<<

T I Geology and Ore Deposits of the i \ Antonio dos Santos, Congo Soco

4 \ and Conceicao do Rio Acima 4

> [ Quadrangles, , s>

( I GEOLOGICAL SURVEY PROFESSIONAL PAPER 341-1 ! i, Prepared in cooperation with the Departamento | Nacional da Producao Mineral of Brazil under ! the auspices of the Agency for International Development of the United States Department ,, of State Geology and Ore Deposits of the Antonio dos Santos, Congo Soco

and Conceicao-^ do Rio Acima Quadrangles, Minas Gerais, Brazil By SAMUEL L. MOORE

GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL

GEOLOGICAL SURVEY PROFESSIONAL PAPER 341-1

Prepared in cooperation with the Departamento Nacional da Producao Mineral of Brazil under the auspices of the Agency for International Development of the United States Department of State

UNITED STATES GOVERNMENT PRINTING OFFICE, WASHINGTON : 1969 UNITED STATES DEPARTMENT OF THE INTERIOR STEWART L. UDALL, Secretary

GEOLOGICAL SURVEY William T. Pecora, Director

For sale by the Superintendent of Documents, U.S. Government Printing Office Washington, D.C. 20402 CONTENTS

Page Abstract______II Structural geology Continued Introduction. ______2 Faults Continued Stratigraphy. ______4 Fundao .. _ __ __. 131 Rio das Velhas Series______4 Cambotas fault. ______31 Group (undivided). 4 High-angle thrust faults.. 32 Maquine" Group (undivided) __. 8 Cambotas homocline...... --. 33 Tamandud, Group (undivided). 10 Summary ______33 Cambotas Quartzite______. 10 Iron deposits.------.-..----. 35 Minas Series.______14 High-grade hard ore. 36 Caraga Group______. 14 Piao prospect. ... . 36 Formation. ._____. 14 Cabega de Ferro mine..-. 38 Batatal Formation. ______16 Santo Antonio mine. 39 Group______16 High-grade soft hematite ore. . 39 Caue1 Itabirite______. 16 . ______40 Gandarela Formation. 19 Canga ______41 Piracicaba Group______. 22 Congo S6co mine______-. 41 Cercadinho Formation..... 22 Barra Mansa mine______. 42 Sabard, Formation. ______23 Cabral mine ______42 Surficial deposits______24 Mina da Trindade. ______42 Canga.______24 Fura Olho mine. ______42 Lateritic soils.______25 Ponte da Paixao mine.... 42 Alluvium. ______25 Other mineral resources. ..___. 43 Talus.______25 deposits...... ----- 43 Igneous rocks.______25 Lufs Soares mine. _____-. 43 Granitic and granodioritic gneiss_. 26 Camard, mine.. ___-_.. 43 Mafic intrusive rocks______27 Congo So6o (gold) mine.. 44 veins______27 Manganese deposits. ______45 Metamorphism. ______28 Capim Gordura mine 46 Structural geology.____---__-______-_. 29 Lag6a das Antas mine..-. 46 Folds______._-______.___.__. 29 Dolomite. ______-__-__------46 Capivara syncline.______29 Ferruginous bauxite. ______46 Conceigao anticline.______29 47 Gandarela syncline. ______30 References cited. 48 Faults______. 30 Index. ______49 Caraga fault.______30

ILLUSTRATIONS

[Plates are in pocket] PLATE 1. Geologic map of the south half of the Ant6nio dos Santos quadrangle. 2. Geologic map of the Congo S6co quadrangle. 3. Geologic map of the Conceigao do Rio Acima quadrangle. 4. Geologic sections of the Congo S6co and Conceigao do Rio Acima quadrangles. FIGTTBE 1. Index map showing location of the Ant6nio dos Santos, Congo S6co, and Conceigao do Rio Acima quadrangles. _ 12 2. Diagram showing correlation of the Tamandud, Group with the Nova Lima Group ______11 3. Cross sections showing inferred structural development in the Congo S6co and Conceigao do Rio Acima quad­ rangles based on section B-B'-B", plate 4______.____-----_-__------34 4. Geologic map and section of the Piao prospect...______-____---_--_-_--_------37 in IV CONTENTS TABLES

Page TABLE 1. Correlation chart of the principal stratigraphic classifications of Precambrian rocks in the Quadrildtero Ferrffero 15 2. Analyses of fresh dolomite and weathered dolomite from the Socorro quarry, Gongo S6co quadrangle...______21 3. Analyses of hard hematite ore from the Cabecja de Ferro mine______-__-__------_-----_------39 4. Analyses of canga ore and enriched itabirite ore from the Gongo S6co mine______42 5. Annual production from the Gongo S6co iron mine____---_-_------42 6. Tonnage and analyses of canga and enriched itabirite ore from the Cabral mine______-___-___---_-----_---_- 42 7. Analyses of manganiferous laterite from Lag6a das Antas mine and manganese ore from Capim Gordura mine___ 47 8. Analyses of dolomite from the Gongo S6co quadrangle.__-__-___-____----_--___------_------47 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL

GEOLOGY AND ORE DEPOSITS OF THE ANTONIO DOS SANTOS, CONGO S6CO, AND CONCEigXo DO RIO ACIMA QUADRANGLES, MINAS GERAIS, BRAZIL

By SAMUEL L. MOORE

ABSTRACT igneous rocks. dikes and veins occur only in contact metanaoxphic zones along the contact with the granitic The AntOnio dos Santos, Gongo S6eo, and Conceigao do Rio Aciina quadrangles lie in the east-central part of the Quad- and are hydrothermal in origin. The Rio das Velhas Series was deformed during at least one rilatero Ferrlfero, or "iron quadrangle," of Minas Gerais, Brazil. period of before the deposition of the Minas Series. The area is part of the Brazilian and is underlain by After the deposition of the Minas Series, both rock series were folded and faulted Precambrian metasediinenitary and igneous broken by high- and low-angles thrust faults and folded into rocks. The metasedimentary rocks have been divided into two northeast-trending folds that are overturned to the northwest. series: the lower, Rio das Velhas Series and the Minas Series, Evidence of the trends of pre-Minas deformation has been separated by a discordant contact. Granitic gneiss intrudes the largely destroyed by post-Minas orogeny, and both series of Rio das Velbas Series, and gabbro and diabase intrude rocks rocks have been metamorphosed to the greenschist facies. of both series. Two major types of occur in the area: one is high- The Rio das Velhas Series has been divided into three units. grade hard and soft hematite ore in the Itabira Group, and the The lower, Nova Lima Group consists of schist, phyllite, and other is surficial canga ore of intermediate grade. iron-formation. The middle, Maquin6 Group consists of quart- The high-grade hematite ore contains 66-69 percent iron and zite, conglomeratic quartzite, and quartz-serlcite schist. The occurs in the iron-formation of the Caue Itabirite and in dolo­ upper, Oamibotas Quartzite consists dominantly of quartzite, mite of the Gandarela Formation. The high-grade hematite ore and sericitic quartzite with a few layers of iron-formation con­ in the Gandarela Formation occurs along the axial zone of a glomerate. All units are separated by unconformable, fault, or subsidiary . The crosscutting relationships of the high-grade discordant contacts. hematite ores to the laminations of the iron-formation and The Minas Series also has 'been divided into three groups. bedding planes of the dolomite show these ores to be replacement The lower Caraga Group is composed of the Moeda Formation, bodies rather than sedimentary in origin. It is believed that they consisting of quartzite, conglomerate, and quartz-sericite schist, were formed by iron-bearing fluids that leached iron from the and the overlying Batatal Formation, consisting of phyllite and surrounding itabirite and then replaced the quartz and car­ argillite. The overlying Itabira Group consists of CauS Ifcabi- bonate of the Gau6 Itabirite and dolomite of the Gandarela rite overlain by the Gandarela Formation. The Caug Itabirite Formation. Textures of the high-grade hematite range from hard is made up dominantly of laminated iron-formation, consisting massive ore through schistose, platy ore to soft, powdery ore. mostly of dolomitic itabirite, siliceous itabirite, and lenses of Most ore bodies contain all three textures. The schistose and dolomite. It contains most of the iron deposits. The Gandarela soft ores are believed to be derived from the weathering and Formation is composed of high-magnesian dolomite interlay- disaggregation of the hard ore. The high-grade hematite ore ered with some dolomitic itabirite and phyllite. Disconfonnably contains minor amounts of chrysotile, quartz, and phosphorus- above the Itabira Group is the Piracicaba Group, composed of bearing gangue minerals. the Cercadinho Formation, consisting of ferruginous quartzite, Surficial canga ore contains 50-60 percent iron and occurs as quartzite, quartz-sericite schist, and phyllite, and the overlying a 1- to 10-rneter-thick layer of conglomerate-breccia composed Sabara Formation, consisting of schist, phyllite, and feldspathic dominantly of fragments of iron-formation cemented by limonite. quartzite. In the mapped area the upper part of the Sahara The canga ore, derived from the weathering of iron-formation, Formation has been removed by erosion. was deposited on a Tertiary erosional surface. The canga ores Only one granitic rock was recognized in the mapped area. range in texture from conglomerate-breccia, the dominant type, It was correlated with the youngest of four that occur to limonite enclosing minor amounts of quartz grains, aluminous in the region (500 million years old on the basis of Ar40/K" ratios and ferruginous laterite, and iron-formation fragments. Canga in biotite). Granitic gneiss cuts across rocks of Rio das Velhas contains widely varying amounts of aluminum- and phosphorus- age and forms part of a regional batholith in the northern part bearing gangue minerals. of the mapped area. Reserves of high-grade hematite ore in the Gongo Soco and Gabbro and diabase, now mostly weathered to lateritic soil Conceigao do Rio Acima quadrangles are estimated to be 2 and saprolite, intrude the Rio das Velhas and the Minas Series. million metric tons above the present surface and about 80,000 These intrusive rocks occur as dikes along faults and joints, as tons per meter of depth. Reserves of intermediate-grade canga plutons, and, rarely, as sills. Quartz veins of several types occur ore are estimated to be 66 million metric tons. Weathered soft along joints and minor faults in the metasedimentary and Cau6 Itabirite averages about 40 percent iron and is a source

II 12 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL of vast tonnages of material that in the future may be benefl- of the Departmento Nacional da Produgao Mineral and ciated for use in iron furnaces. the U.S. Geological Survey. The U.S. Geological Sur­ Dolomite, a potential source of dimension and crushed stone; vey work in this program was done under the auspices clays, suitable for the manufacture of roofing tile, fire brick, and refractory brick; ferruginous bauxite, a potential source of low- of the Agency for International Development of the grade aluminum ore, and minor amounts of manganese occur in U.S. Department of State. the mapped area. Lode and placer deposits of gold were mined The Conceigao do Rio Acima, Gongo Soco, and An­ during the 18th and early part of the 19th centuries, but they tonio dos Santos quadrangles are in the east-central part were abandoned because of diminishing ore grade and depletion of the Quadrilatero Ferrifero (fig. 1), about 35 kilo­ of ore. meters east of . The quadrangles are INTRODUCTION accessible from Belo Horizonte by both a highway and It has long been known that the Quadrilatero Fer- the Central do Brasil railroad, which pass through rifero, or iron quadrangle, of central Minas Gerais con­ the central part of the Gongo Soco quadrangle. The tains large tonnages of high-grade iron ore (more than area is traversed by a network of secondary roads and 66 percent iron) as well as other lower grade deposits. trails that lead northward and southward from the Because of this potential, Brazil and the United States highway to the mines, ranches, and quarries. The main in 1946 jointly agreed to make a long-term study of the industries of the area are of iron, manganese, regional geology and an evaluation of the iron-ore de­ and dolomite and the manufacture of charcoal for the posits. This study was carried out jointly by geologists iron smelters.

45' 70' 62° 54' 46' 44°00'

20° 00' LOCATION OF PROJECT MAPPING BY U.S. GEOLOGICAL SURVEY AND DEPARTAMENTO NAC10NAL DA PRODUCAO MINERAL

15

16

QUADRANGLE NAMES 1 Itabira 24 Gandarela 2 Santa Luzia 25 Conceicao do Rio Acima 3 Serra da Piedade 26 4 Ant6nio dos Santos 27 Lagoa Grande 5 Coeais 28 6 Sao Gor.falo 29 Rio de Pedras 7 Monlevade 30 Capanema 8 Beio Horizonte 31 Santa Rita Durao 9 Nova Lima 32 Marinho da Serra lOCaete 33 Ba$ao 11 Gongo S6co 34 Caehoeira do Campo 12 Santa Barbara 35 Sao Bartolomeu 13 Floralia 36 Antonio Pereira 14 37 Casa de Pedra 15 Serra Azul 38 S5o Juliao 16 Itatiaiuc.u 39 Dom Bosco 30' 17 Igarape 40 Ouro Branco 18 Sousas 41 Ouro Pr&to 19 Fe'eho do Funil 42 Santa Rita do 20 43 Mariana 21 Ibiritt 44 Rio das Bandeiras 22 Macacos 45 23 Rio Acima 46

10 20 MILES _]_ I 10 20 KILOMETERS I I 20"45' FIGURE 1. Location of the AntSnio dos Santos, Gongo Sftco, and Conceigao do Rio Acima quadrangles. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 13

In the east-central part of the Quadrilatero Ferrlfero along the Central do Brasil railroad. The Gongo Soco are four major mountain ranges: the Serra do Caraga, open-pit iron mine is operated by Companhia Ferro the Serra das Cambotas, the Serra Geral, and the Serra Brasileiro, Usina Gorceix, which is mining the surficial da Pedra Formosa. The Serra Geral and the Serra da canga iron ores. The Cabral and Trindade mines, now Pedra Formosa trend northeast and extend diagonally closed, were operated by Companhia Siderurgica Belgo- across the area between the north-trending Serra das Mineira from 1948 to 1954. The CabeQa de Ferro iron Cambotas on the northeast and the Serra do Caraga on mine, near the east boundary of the Gongo Soco quad­ the southeast. The maximum relief in the area is about rangle, is operated by the Companhia Brasileira da 1,180 meters. The highest point in the area, 1,911 meters, Usinas Metalurgicas. is in the Serra do Maquine; the lowest point about 730 Dolomite is produced from many small quarries in meters, is in the valley of the Rio da Conceigao. The the Gongo Soco quadrangle for use as fluxing material. Rio da Conceigao flows northeast between the Serra do Manganese has been produced in small quantities in the Caraga on the southeast and the Serra da Pedra For­ Gongo Soco quadrangle. mosa on the northwest. The Rio Socorro also flows The geology and ore deposits of the Quadrilatero northeast, between the Serra Geral on the northwest Ferrlfero have been studied for more than 130 years, and the Serra da Pedra Formosa on the southeast. These and a large bibliography has accumulated. Most of the two rivers flow into the Rio Doce and drain about 300 studies have been of individual ore deposits. One of the square kilometers of the area that lies south of the Serra earliest studies of the regional geology of the Quad- Geral. The low rolling country north of the Serra Geral riMtero Ferrifero was made by Eschwege (1883). and west of the Serra das Cambotas is drained by a Later regional studies of the geology of the Quadrild- series of small streams that flow northward into the tero Ferrifero were made by Derby (1906), Harder Rio das Velhas. and Chamberlin (1915), Freyberg (1932), and Gui- The region has a subtropical climate; seasonal tem­ maraes (1951). Dorr and Bafbosa (1963), Guild (1957), peratures range from about 0°C to about 37°C, and Gair (1962), and Pomerene (1964) discussed the ore de­ the average daytime temperature is about 17 °C during posits of the region. the winter and about 23° C during the summer. Rainfall Geologic mapping for the present investigations was is seasonal and occurs mostly from November to April, begun in June 1955 and was finished in October 1957. when torrential rains are common. From April to No­ Rocks of the Minas Series were mapped in greater de­ vember there is little precipitation, and the sky is gen­ tail than other rocks because they contain most of the erally clear. The average annual rainfall is about iron-ore deposits and because the outcrops are more 1,500 mm. abundant. Pre-Minas rocks are deeply weathered, crop The valleys are commonly covered with a thick growth out sparsely, and are of slight economic interest. of trees, vines, and ferns; the intermediate elevations Mapping data were recorded in the field on aerial on the flanks of the serras are generally timbered with photographs at a scale of 1:25,000 and later transferred a scrubby growth of trees; and the higher areas are to topographic base maps at a scale of 1:20,000. To covered with grass. Most of the thickly wooded areas allow location of points that do not have formal names, have been cut over several times for charcoal, and virgin a coordinate grid system was used on each quadrangle mato (jungle) remains only on the steeper slopes where map (pis. 1-3). In this system, the 0 point, or the origin, timbering is difficult and uneconomical. is at the southwest corner of each map, and tick marks Gold was the first metal sought by prospectors and on the sides of the map indicate 1,000-meter intervals miners in the central part of Minas Gerais. Portuguese away from this point. settlers prospected widely in the valleys of the Rio da The cooperation, both in the field and in the labora­ Conceigao and the Rio Socorro and operated large tory, of all members of the Departmento Nacional da placer mines during the early part of the 19th century. Produgao Mineral is gratefully acknowledged. The Underground mining for gold was started in 1824, writer especially thanks Dr. Jose Alves, Chief of the when the Imperial Mining Co., a British firm, pur­ Belo Horizonte office, for his cooperation in assigning chased the Gongo Soco mine. This mine was in opera­ field personnel and providing laboratory facilities dur­ tion until 1856, when it was closed because of depletion ing all phases of the study, and Dr. Cassio Pinto, Chem­ of ore. ist of the D.N.P.M. laboratory in Belo Horizonte, who Open-pit iron mines have been operated since 1935 made most of the analyses of ores. Mr. Charles Wurth in the Gongo Soco quadrangle. The Barra Mansa mine, of the Ferro Brasileiro Usina Gorceix and Dr. Fran­ the Trindade mine, the Gongo Soco iron mine, and the cisco Jose Pinto de Souza of the Companhia Brasileira Cabral mine are on the south slope of the Serra Geral, da Usinas Metalurgicas provided company files from 14 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL which tonnage and grade of ore were obtained. Dr. Luis the basal formation of the Tamandua Group defined by Verano furnished lodging at Casa de Hospedes in Simmons and Maxwell (1961). Barao de Cooais and at Fazenda Capivara. The field mapping was carried out under the super­ NOVA LIMA GROUP (UNDIVIDED) vision of J. V. N. Dorr 2d, chief of party in Brazil, and DISTRIBUTION W. D. Johnston, Jr., U.S. Geological Survey, Wash­ Schist, phyllite, and iron-formation of the Nova Lima ington, D.C. G. A. Rynearson, U.S. Geological Survey, Group have been mapped in a northeast-trending belt supervised the initial phases of the work. J. B. Pom- across the Conceigao do Rio Acima quadrangle and the erene, a fellow worker, provided much good advice southeastern part of the Gongo Soco quadrangle (pis. 2, and assistance during the first 2 months of fieldwork. 3). This belt of Nova Lima rocks is overlain by the basal Jose Eoberto da Silva assisted the writer in the field, formations of the Minas Series on the northwest and served as chauffeur, and was a very pleasant companion by schist and quartzite of the Maquine Group on the during the fieldwork. southeast. A band of quartz-chlorite schist, iron-formation, and STRATIGRAPHY some dolomitic phyllite has been mapped across the The Antonio dos Santos, Gongo Soco, and Conceigao north-central part of the Gongo Soco quadrangle (pi. do Rio Acima quadrangles are underlain by regionally 2). On the south these rocks are overlain by the basal metamorphosed sedimentary and igneous rocks of Pre- formations of the Minas Series, and on the north the cambrian age (pis. 1-3). The metasedimentary rocks rocks are intruded by granodiorite gneiss of the Caete are schist, phyllite, quartzite, dolomite, and itabirite. Complex, from the west boundary of the quadrangle to The igneous rocks are granitic gneiss, diabase, and near N. 7,800, E. 6,800 (pi. 2). From this locality to the gabbro. east boundary of the quadrangle, rocks of the Nova The metasedimentary rocks of the quadrangles are Lima Group overlie and underlie Cambotas Quartzite divided into two series (Dorr and others, 1957) : an along the south and north slopes of the Serra Geral older, Rio das Velhas Series, and a younger, Minas (pi. 2). Series. All the sedimentary rocks of the Quadrilatero The stratigraphic position of this belt of schist and Ferrifero were defined as Minas Series by Derby (1906, iron formation in the column is uncertain. West of the p. 396). Harder and Chamberlin (1915) included most Gongo Soco quadrangle these rocks have been mapped of the metasedimentary rocks of the region in their in the Caete quadrangle as Nova Lima Group (N. L. Minas Series (table 1). Barbosa (1954), Guimaraes Alves, oral commun., 1955). In the Nova Lima quad­ (1931, p. 8), Oliveira (1956), and Dorr, Gair, Pomerene, rangle, Gair (1962, p. 16-26) included rocks of similar and Rynearson (1957) divided Derby's original Minas lithology in the Nova Lima Group. After the comple­ Series into three series (table 1). They named the upper tion of this study, Simmons and Maxwell (1961) meas­ series Itacolomi (Guimaraes, 1931) and the middle ured a section in the northeastern part of the Gongo series Minas. The lower series was named the Barba- Soco quadrangle and the eastern part of the Santa Bar­ cena Series by Barbosa (1954), the pre-Minas Series by bara quadrangle. They assigned this belt of schist and Oliveira (1956), and the Rio das Velhas Series by Dorr, iron-formation to the upper part of their Tamandua Gair, Pomerene, and Rynearson (1957). Barbosa's Bar- Group and divided these rocks into three unnamed bacena Series included some rocks of known post-Minas units. Thrust faults along the north boundary of this age, and he indicated that his Series may belt of rocks further complicate positioning these rocks correlate with the rocks that are now called Rio das in the stratigraphic column and determining their age. Velhas Series (1954, p. 21). However, this belt of rocks was traced westward from the Santa Barbara quadrangle across the Gongo Soco EIO DAS VELHAS SEEIES quadrangle into rocks of unquestioned Nova Lima age Rocks of the Kio das Velhas Series (Dorr and others, in the Caete quadrangle and they are therefore included 1957, p. 11-22) have been divided into three units: the in the Nova Lima Group. Nova Lima Group (undivided), at the base, consisting of schist, phyllite, and iron-formation; the Maquine LITHOLOGY Group (undivided), consisting of quartzite, conglom­ In most places where the schist, phyllite, and iron- erate, quartz-chlorite schist, conglomerate schist, and formation are exposed, they are deeply weathered; out­ phyllite; and the Cambotas Quartzite, consisting of crops of fresh rock are very sparse. The schist and pebble conglomerates, quartzite, sericitic quartzite, and phyllite weather to a surficial layer of brownish-red to minor amounts of phyllite. The Cambotas Quartzite is dark-red lateritic soil mantling a deeply weathered ANTONIO DOS SANTOS, GONGO SOCO, AND CONCBigAO DO RIO ACIMA QUADRANGLES I 5

TABLE 1. Correlation chart of the principal stratigraphic classifications of Precambrian rocks in the Quadrildtero Ferrifero

U.S. Geological Survey and the Departamento Harder and Chamberlain (1915) Barbosa (1954) Oliveira (1956) Nacional da Producao Mineral, in 1961, combined with Simmons and Maxwell (1961)

ItacolomJquartzite ItacolomiSeries *c.2

M

Sahara Formation; phyllite, gray- wacke, subgraywacke, and meta- volcanic rocks

Barreiro Formation; schist, phyllite, PiracicabaFormation FormationPiracicaba o! and graphitic phyllite PiracicabaFormation! Phyllite, quartzite, Quartzite, phyllite, Tabooes Quartzite; quartzite Schist, quartzite, and dolomite, and iron- ferruginous quartz­ dolomite formation (itabirite) ite, and graphitic 1 Fficho do Funil Formation; dolomitic phyllite S phyllite, argillaceous dolomite, phyllite, and ferruginous quartzite

Cercadinho Formation; quartzite, MinasSeries MinasSeries 5S2* ferruginous quartzite

M Itabira Formation Dolomite, iron-forma­ Itabira Formation tion (itabirite), and Itabirite and dolomite ItabiraGroup Gandarela Formation; dolomite, phyllite 1 iron-formation, and phyllite Minasseries ItabiraFormation Cau6 Itabirite; itabirite (oxide-facies Batatal Iron-formation Formation iron-formation) (itabirite) Phyllite CaragaFormation2 Batatal Formation; phyllite, and minor amount of schist Quartzite and CaracaGroup conglomerate "3« CaragaFormation Quartzite and Moeda Formation; quartzite, sericite OS'S pqco Schist conglomerate quartzite, and conglomerate

Quartz-chlorite schist TamanduaGroup5 Ferruginous dolomitic quartzite

Quartz-chlorite schist

« Cambotas Quartzite; quartzite and o> SeriesBarbacena4 RiodasVelhasSeries sericitic quartzite 1ta Quartz and mica M Quartzite, conglomer­ Greenschist, talc schist, gneiss, phyl­ Upper part; quaitzite conglomerate 8 VD o1 ate, and quartz and schist, gneiss, and lite, graywacke, iron- a and schist 03 § mica schist granodiorite formation, quartz­ *5*1 C3 ?a ite, and dolomite Lower part; quartzite, graywacke, 0 quartz schist, and phyllite

NovaLima Phyllite, greenschist (metasedimen- Group tary and metavolcanic rocks), iron- formation (carbonate), quartzite, and graywacke

Basalcomplex Mantiqueira .§ Probable 2 a Gneiss, amphibolite, Series .D'-g and Gneiss s ^ Gneiss Gneiss and granitic rock

' Includes the Batatal of Harder and Chamberlain (1915). 2 Includes Rio das Velhas rocks classified as Maquine1 Group in the U.S. Geological Survey-Departamento Nacional de Producao Mineral classification. * Includes part of the Rio das Velhas Series in the U.S. Geological Survey-Departamento Nacional de Producao Mineral classification. 4 Includes some rock of the Itacolomi Series and post-Minas granitic rocks. ' Includes Caraca quartzite of Harder and Chamberlain (1915) in eastern part of the Quadrilatero Ferrifero. «Includes igneous rocks of at least three ages ranging from pre-Rio das Velhas to post-Minas. 309-760 68 2 16 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL saprolite,1 and most of the rock descriptions are those of poorly exposed in most localities, but good exposures of weathered material. Saprolite crop out in the railroad cuts about 1,600 me­ Schist, phyllite, iron-formation, and graphitic phyl- ters northeast of the Camara gold mine and in roadcuts lite and schist of unquestioned Nova Lima age crop out along the old road leading southward from Joao in readouts and in valleys in the Conceigao do Eio Vasconcelos to the old abandoned Gongo Soco gold Acima quadrangle (pi. 3). Schist weathers to light-red- mine. The quartz-chlorite schists are reddish brown dish-brown to dark-red friable material. Granular and poorly laminated and consist of darker chlorite- quartz and bleached mica are the only recognizable min­ rich bands alternating with bands of granular quartz erals. Phyllite weathers to light-red and purple or and chlorite interlayered with thin dolomitic iron-for­ brownish-red and dark-gray earthy material consisting mation that is oxidized to limonite and fine granular of bleached micaceous minerals and arenaceous material quartz. embedded in a clay matrix. A band of light-gray and silvery-gray quartz-sericite Thin layers of dark-gray graphitic schist and schist and muscovite schist unconformably underlies phyllite are exposed in the belt of Nova Lima rocks the Cambotas Quartzite along the north slopes of the underlying the Conceigao valley (pis. 2, 3). They are Serra Geral and along the southwest escarpment of the interlayered with dark-red quartz-chlorite schist, red­ Serra das Cambotas. On the north and the east it is dish-brown phyllite, and thin iron-formations in road- truncated by granodiorite gneiss. This narrow belt of cuts. The graphitic schist and phyllite layers are 1-3 Nova Lima rocks consists of muscovite-quartz schist meters thick and weather to finely laminated fragments and a few interbeds of reddish-gray conglomeratic and dark-gray to black residual soils. In places where phyllite. The lower beds of schist adjacent to the grano­ the graphitic schists are not too severely weathered, the diorite gneiss of the Caete Complex contain metacrysts laminae are either crenulated or have a splintery of feldspar and some disseminated specular hematite, texture. and in places along the Caete-Barao de Cocais highway A belt of quartz-sericite schist, quartz-chlorite schist, they are migmatized over a zone 10-100 meters wide. iron-formation, and dolomitic phyllite of the Nova Lima Group crops out in roadcuts and along trails on IRON-FORMATION the south slopes of the Serra Geral in the north-central Two prominent layers of iron-formation are exposed part of the Gongo Soco quadrangle (pi. 2). Ked and in the northeastern part the Conceigao do Eio Acima gray fine-grained quartz-chlorite schist and quartz- quadrangle and the southeastern part of the Gongo sericite schist compose the upper beds of this belt of Soco quadrangle (pis. 2, 3). They range from a few Nova Lima rocks. They underlie the Caue Itabirite of meters to about 150 meters in thickness. Good exposures the Minas Series from the Gongo Soco iron mine to the of the iron-formations can be seen along the headwaters east boundary of the quadrangle (pi. 2). Conformably of the Corrego do Sitio and in a roadcut about 1,000 beneath the schist is 40-60 meters of interlayered dolo­ meters north of the village of Campo Grande. Else­ mitic iron-formation and dolomitic phyllite. Conform­ where in Conceigao valley, a series of thin lenses of iron- ably below the iron-formation and dolomitic phyllite formation from 0 to 10 meters thick is sporadically in­ is about 15-30 meters of interbedded reddish-brown terlayered with schist and graphitic schist of the Nova ferruginous quartzite and white schistose quartzite. Lima Group. In the north-central part of the Gongo This layer of quartzite is more resistant to erosion and Soco quadrangle (pi. 2), a prominent layer of dolomitic crops out at many localities 'along the south slopes of iron-formation interlayered with phyllite was mapped the Serra Geral. It is in contact with the Caue Itabirite from the west boundary of the quadrangle at N. 5,800 of the Minas Series near the open pits of the Camara to N. 5,500, E. 4,200, and from N. 8,500, E. 12,000 to the gold mine and in a railroad cut 300 meters west of the east boundary at N. 9,050, E. 13,100. Thin lenses of iron- Gongo Soco iron mine. Elsewhere in the Gongo Soco formation occur at many levels in the schists of the quadrangle it occurs 0-150 meters beneath the basal Nova Lima Group along the south slopes of the Serra formations of the Minas Series. Geral. Reddish-brown quartz-chlorite schists interlayered The thicker layers of iron-formation in the Conceicao with thin iron-formation occur beneath the schistose valley are interlayered with ferruginous quartz-chlo­ and ferruginous quartzite and above the Cambotas rite schist and graphitic phyllite. The upper and lower Quartzite. These schists are deeply weathered and contacts of well-laminated layers of dolomitic iron- 1 Saprolite is an earthy material that has been formed essentially in formation are gradational with the schist and phyllite place as the result of weathering of bedrock. Structures and textures of of the Nova Lima Group over a width of about 2-5 the rocks are usually partly or totally retained, but the mineralogy is completely changed by weathering. meters. The rocks at the contacts grade from well- ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 17 laminated dolomitic iron-formation through a zone of along strike over a few tens of meters into ferruginous indistinct laminae showing considerable mixing of the quartz-chlorite schist.

iron-bearing mineral with quartz and dolomitic carbon­ THICKNESS AND STRATIGRAPHIC RELATIONS ate, into ferruginous schist and phyllite. At N. 900, E. 12,800 in the southeast corner of the The thickness of the Nova Lima rocks is not known in Gongo Soco quadrangle (pi. 2), a 150-meter sequence of this area or elsewhere in the Quadrilatero Ferrifero be­ iron-formation, ferruginous schist, and graphitic schist cause the base of the group has not been recognized or is consists of 30 meters of low-grade poorly laminated not exposed. A thickness of at least 2,000 meters is in­ iron-formation interlayered with dark-gray graphitic dicated in the central part of the Conceigao do Eio schist overlain by 20 meters of dark-red ferruginous Acima quadrangle (pi. 3), but isoclinal folds and faults quartz-chlorite schist. Conformably above the schist is may duplicate beds. 90-100 meters of well-laminated dolomitic iron-forma­ Eocks of the Nova Lima Group underlie the basal tion. The low-grade iron-formation and the well-lami­ formations of the Minas Series with slight to moderate nated iron-formation in most places are weathered to a discordance in the northwestern part of the Conceic,ao granular mass of hematite, fine powdery talc, limonite, do Eio Acima and the southeastern part of the Gongo Soco quadrangle. Beds of the overlying Minas Series and granular quartz that crumbles with the slightest blow. Fibrous tremolite occurs in the light-colored and the underlying Nova Lima Group are truncated by this discordant contact. A strong angular discordance bands, and a few thin laminae of manganese oxide par­ allel the laminations in some layers. Quartz veins cut between schist of the Nova Lima Group and the overly­ across the laminations in many layers. All the iron-for­ ing quartzite of the Moeda Formation of the Minas mation is dolomitic and strongly magnetitic. A thin Series can be seen near the west boundary of the Con- ceigao do Eio Acima quadrangle at N. 5,300, E. 1,400 section of weathered but indurated iron-formation from (pi. 3). Here a difference of 40° in strike and 40° in dip a small open pit at N. 825, E. 12,800 (pi. 2), shows the was found between the schist and the overlying quartz­ iron-bearing laminae to be made up of hematite and partially hydrated to limonite, intercrystal- ite. The contact is well exposed in the Corrego da lized with minor amounts of quartz and dolomitic car­ Gandarela. The basal 2 meters of quartzite and conglom­ bonate. The light-colored laminae consist of a mosaic of erate of the Moeda Formation is strongly foliated, and quartz pebbles and cobbles are stretched and sheared. quartz and dolomitic carbonate enclosing a few grains of hematite and magnetite. Hematite and magnetite grains Northeastward from the Corrego da Gandarela the con­ tact is well exposed in several corregos, and the dis­ range from about 0.01 mm to 0.4 mm across and have cordant contact separating Minas and pre-Minas rocks a median size of 0.15 mm; the quartz and carbonate obliquely truncates and cuts out quartzite of the Moeda grains range from 0.02 to about 0.25 mm across and Formation .and phyllite of the Batatal Formation near have a median size of about 0.10 mm; the thickness of N. 8,900, E. 3,700 (pi. 3). At N. 10,000, E. 4,400 (pi. 3), the laminae ranges from 1 mm to about 2 cm. along the south slopes of the Serra da Pedra Formosa, Iron-formations in the belt of Nova Lima rocks ex­ a thin, highly sheared sericitic quartzite conglomerate posed along the south slopes of the Serra Geral in the occurs below the phyllite of the Batatal Formation and Gongo Soco quadrangle are interlayered with dolomitic above the schist of the Nova Lima Group. From N. phyllite and occur in a layer ranging from about 30 to 11,000, E. 5,100 (pi. 3), in the northwestern part of the 60 meters in thickness. Individual layers of well-lami­ Conceigao do Eio Acima quadrangle to the vicinity of nated dolomitic iron-formation range from a few centi­ Lagoa das Antas in the southeastern part of the Gongo meters to as much as 10 meters in thickness. The darker Soco quadrangle the contact between the thin basal colored iron-rich laminae consist of hematite and mag­ quartzite of the Minas Series and the Nova Lima Group netite and alternate with lighter-colored laminae of is poorly exposed. In most places it is covered by surficial deposits of canga, talus, and lateritic soil. Little granular quartz and dolomitic carbonate. Most of the is known about the contact except that the quartzite iron-rich laminae are weathered to very fine grained above and the schist below apparently dip and strike red limonite, and the quartz and dolomitic carbonate nearly parallel to it. About 700 meters east of Lagoa das laminae are weathered to talc and iron-stained granular Antas, well-exposured Moeda Formation stratigraphi- quartz. The thinner lenses of iron-formation are gen­ cally overlies schists of the Nova Lima Group with erally weathered to limonite laminae alternating with slight disconformity. Schist of the Nova Lima Group very fine powdery granular quartz and in most places stratigraphically underlies Caue Itabirite at N. 4,950, are either interlayered with thin graphitic schist and E. 12,050 (pi. 2), near the east boundary of the Gongo phyllite or ferruginous schist and phyllite. They grade Soco quadrangle, whereas 1,600 meters to the southwest 18 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL and 1,000 meters to the northeast basal Minas quartzite to moderate discordance with the underlying Caue of the Moeda Formation stratigraphically overlies Itabirite of the Minas Series. schist of the Nova Lima Group. The Caue Itabirite 100 The stratigraphic position of the formations of the meters southeast of the Ponte da Paixao mine is severely Minas Series is well known in the western part of the distorted by drag folds 'and is cut by closely spaced area (pis. 2, 3), and the absence of the basal Moeda shear planes about 20 meters from the contact. About Formation, the Batat'al Formation, and part of the 200 meters southeast of the Fura Ohio mine, phyllite Caue Itabirite from the section along the eastern part of and lenses of sericitic quartzite conglomerate about 15 the Gandarela syncline could be due to either nondeposi- meters stratigraphically above the contact are broken tion or fault displacement along the contact separating and sheared, and lenses of the quartzite are sliced one the rocks of the Minas Series from the rocks of the over the other. The contact is not exposed, but it is Nova Lima Group. The stratigraphic and structural evident that the basal formations of the Minas Series discordances along this contact are further discussed are distorted by a fault zone. on pages 133-135. Along the south slopes of the Serra Geral, from N. In the southeastern part of the Conceigao do Bio 4,600 on the west boundary to near N. 5,100, E. 4.850 (pi. Acima quadrangle, schist and quartzite of the Maquine 2), schist of the Nova Lima Group underlies sericitic Group overlie schist of the Nova Lima Group with ap­ quartz schist of the Moeda Formation ,with apparent parent discordance. A strong discordance in dip of conformity. However, about 900 meters to the northeast, schist above and below the contact can be seen on a ridge near the Camara mine workings, the basal formations about 2,000 meters southeast of Serra Kedonda (p. 3). of the Minas Series are missing from the section, and The contact is covered by lateritic soil, but about 20 Caue Itabirite of the Minas Series overlies ferruginous meters above and below the contact the schists show quartzite and schistose quartzite of the Nova Lima difference in dip of about 30°. Elsewhere along this con­ Group. The contact between the itabirite and the quartz­ tact, only slight to moderate angular discordance is in­ ite is well exposed in open pits ,and in a partly caved dicated ; however, the contact is covered or poorly ex­ inclined shaft and consists of 50-70 em of sheared and posed, and it is not known whether it is an unconformity silicified ferruginous quartzite that is overlain by or a fault. highly crenulated itabirite and underlain by ferrugi­ Schist and phyllite of the Nova Lima Group are cut nous quartzite broken by numerous tear fractures. The by granodiorite gneiss of the Caete Complex in the cen­ contact dips 45° S., and the highly distorted lamina­ tral and northeastern parts of the Gongo Soco quad­ tions of the itabirite are dragged into the contact and rangle. The chlorite-quartz schist in the central part of truncated by the . the Gongo Soco quadrangle forms a sharp contact with Caue Itabirite overlying ferruginous quartzite is well granodiorite in readouts and railroad cuts. In the north­ exposed in railroad cuts 300-500 meters northwest of eastern part of the Gongo Soco quadrangle the sericite- the Gongo Soco iron mine at N. 6,600, E. 7,100 (pi. 2). schist and muscovite-quartz schist of the Nova Lima Here the laminations in the itabirite above the contact Group are migmatized and form a contact zone of are severely distorted by tight drag folds ranging in hybrid rocks containing numerous metacrysts of feld­ amplitude from a few centimeters to a few meters in a spar and poorly developed gneissic layering that ranges zone 10-15 meters wide. The laminations are dragged from a few meters to about 50 meters outward from into the contact and truncated by it. The drag folds fade out upward from the contact through a zone of about granodiorite gneiss. 10-15 meters into undistorted laminations; that are con­ MAQUINE GROUP (UNDIVIDED) formable with the strike and dip of the northwest limb The Maquine Group has been divided into two forma­ of the Gandarela syncline. The ferruginous and schist­ tions: an older, Palmital Formation (J. E. O'Kourke, ose quartzites below the contact strike and dip nearly unpub. data), comprised of quartz-sericite schist, con­ parallel to the contact ,and are broken by tear fractures. glomeratic schist, subgraywacke, and lenses of sericitic Eastward from the Gongo Soco iron mine to near N. 8,500, E. 12,000 (pi. 2), Caue Itabirite overlies schist quartzite; and a younger, Casa Forte Formation (Gair, of the Nova Lima Group. The itabirite and the schist 1962, p. 31-33), comprised of massive quartzite con­ dip and strike nearly parallel to the contact and are glomerate, sericitic quartzite, and minor amounts of exposed within a few centimeters of one another. East­ phyllite and schist. During the present study, rocks ward from N. 8,500, E. 12,100 (pi. 2), to the east lithologically similar to those described by O'Rourke boundary of the Gongo Soco quadrangle, iron-forma­ and Gair were recognized, but no division could be tion and underlying ferruginous quartzite of the Nova made because of poor exposure and the lack of marker Lima Group diverge from the contact and form a slight beds. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 19

DISTRIBUTION E. 4,400; and along the Corrego do Capivari from N. Eocks of the Maquine Group have been mapped along 200, E. 8,450, to N. 100, E. 8,250, near the south bound­ the north, northwest, and west slopes of the Serra do ary of the Conceigao do Eio Acima quadrangle. Thin Caraga in the southeastern part of the Conceigao do Eio sections from these localities show the rocks to be quartz- Acima quadrangle (pi. 3). Cambotas Quartzite overlies sericite schist, quartz-chlorite schist, metasubgray­ quartz-chlorite schist, conglomeratic schist and phyllite, wacke, and conglomeratic phyllite. The quartz-sericite and subgraywacke from the west boundary at E. 8,300 schists consist of 50-80 percent angular quartz grains to near N. 5,000, E. 9,500, in the southern part of the ranging from 0.4 to 1.0 mm across, and 20-50 percent Conceigao do Eio Acima quadrangle. From the last sericite, chlorite, chloritoid, and minor amounts of py- mentioned locality to the east boundary of the quad­ rite and hematite. Sericite is the predominant micaceous rangle Oambotas Quartzite overlies massive quartzite mineral in most of the schist; it occurs as thin laminae and conglomerate in the upper part of the Maquine of compact plates along lines of quartz grains 'and along Group. Quartz-chlorite schist of the lower part of the boundaries of the quartz grains. Penninite, with its Maquine Group stratigraphically overlies schist of the characteristic ultraviolet interference color, and an un­ Nova Lima Group from the south boundary to the east identified brown chlorite are scattered through the seri­ boundary of the Conceigao do Eio Acima quadrangle cite laminae in most of the schist. A few beds of (pi. 3). pyritized schist, however, contain brown chlorite lami­ nae but only minor amounts of sericite. Chloritoid LITHOLOGY occurs in widely varying amounts in the schist as The rocks of the Maquine Group consist of a belt rosettes and lath-shaped grains. It comprises as much of massively to very thick bedded light-gray to light- as 20 percent of some of the schist, but generally it is brown coarse-grained quartzite, conglomerate, 'and only 1-5 percent of the rock or is absent. Pyrite and sericitic quartzite that is structurally overlain and hematite are sparsely disseminated mostly in the fine­ underlain by brown and reddish-brown thin- to grained schist, but both minerals have been either medium-bedded fine- to medium-grained quartz- partly or almost completely oxidized and hydrated to chlorite schist, conglomeratic quartz-sericite schist, limonite. and a few beds of metagraywacke, and lenses of THICKNESS AND CONTACT RELATIONS quartzite. A thickness of 1,000-1,500 meters is indicated for the The upper quartzite, conglomeratic quartzite, and rocks of the Maquine Group along the northwestern sericitic quartzite are composed of angular quartz grains part of the Serra do Caraga (pi. 3). An undetermined ranging from about 0.5 to 3.0 mm across, and inter­ thickness of the lower beds is cut out of the section by stitial grains ranging from about 0.02 to 0.5 mm across. the Caraga along the northwest escarpment The quartzite contains minor amounts of sericite, chlo- of the Serra do Caraga. Some of the lower schist beds rite, chloritoid, and kyanite. Some quartzite beds con­ may be cut out on the northwest and north at the contact tain about 5-10 percent sericite that is crudely alined in with the underlying schist of the Nova Lima Group, thin discontinuous laminae parallel to the bedding. Cob­ if this contact is a fault. bles and pebbles of vein quartz, quartzite and recrystal- The contact between the Maquine rocks and the over­ lized chert, 10-80 mm in diameter, are randomly lying Cambotas Quartzite is discordant and has been distributed in irregular zones at many levels within the mapped as a thrust fault. This contact is well exposed massive quartzite beds. The conglomeratic zones cut in the Corrego do Capivara at N. 200, E. 8,400 (pi. 3), across bedding planes, and the cobbles and pebbles of near the south boundary of the Conceigao do Eio Acima quartz and quartzite make up less than 10 percent of quadrangle. Here there is a 30° difference in the strike the quartzite. and a 10° difference in the dip between the schist and These coarse-grained quartzites comprise the core of conglomeratic schist of the Maquine Group and the a syncline overturned to the west, northwest, and north. overlying Cambotas Quartzite. The contact consists of They grade upward and downward through thick- 50-80 cm of sheared iron-stained quartz and chlorite; bedded medium-grained brown and reddish-brown it dips about 20° E., strikes north, and cuts obliquely quartz-sericite schist into thin-bedded red and brown across and truncates the massive beds of Cambotas quartz-sericite schist, conglomeratic quartz-sericite Quartzite. The quartzite beds are warped into the fault schist, metasubgraywacke, and a few lenses of quartzite. and are progressively cut out of the section to the south Good exposures of the middle and lower rocks crop out of the Corrego do Capivara, where the Cambotas Quart­ (pi. 3) along the road to the Colegio Caraga road from zite is markedly thinned southward along the west N. 7,700, E. 11,600, to N. 7,200, E. 11,650; near N. 450, escarpment of the Serra do Caraga. The schists of the 110 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL lower part of the Maquine Group are also crosscut by study of the Caete uplift area, correlated the quartzite the fault and cut out of the section southward from the of the Serra das Cambotas with the Caraga Quartzite of Corrego do Capivari. Northwestward from the Corrego Harder and Chamberlin. On a preliminary regional do Capivari along the northwest escarpment of the map (unpub. data, 1959) Dorr tentatively correlated Serra do Caraga, the Caraga fault also cuts across the the quartzite of the Serra das Cambotas with the Moeda lower beds and obliquely truncates the basal schists of Formation (Wallace, 1958). the Maquine Group. Near N. 5,000, E. 9,500 (pi. 3), the Because of a marked structural and stratigraphic dis­ basal schist of the Maquine Group is cut out of the continuity between the quartzite of the Serra das Cam­ section. Near N. 4,700, E. 9,400 (pi. 3), there is a strong botas and the Minas Series at the south end of the Serra angular discordance between the schist and quartzite das Cambotas, previous correlations of the north-trend­ of the Maquine Group and the overlying Cambotas ing belt of quartzite with the basal quartzite of Quartzite. The schist and quartzite strike N. 30° E. and the Minas Series were open to question. The rocks dip 30° SE., and the overlying Cambotas Quartzite of the Minas Series are tightly folded into an strikes N. 10° W. and dips 40° E. Here the fault contact isoclinal syncline that trends N. 70° E., dips about dips about 25° SE. Northward from N. 4,200, E. 9,800, 35° S., and is overturned to the north. In contrast, the to N. 6,200, E. 9,500 (pi. 3), the contact between the belt of quartzite that forms the Serra das Cambotas schist of Maquine age and the Cambotas Quartzite trends north and dips 30°-35° E. The basal quartzite steepens from a dip of about 30° SE. to nearly vertical. of the Minas Series is either missing from the section Along the north escarpment of the Serra do Caraga, or less than 20 meters thick along the east end of the from the area of N. 6,200, E. 9,500 (pi. 3), to the east Gandarela syncline at the south end of the Serra das boundary of the Conceigao do Eio Acima quadrangle, Cambotas in the eastern part of the Gongo Soco quad­ the contact between the quartzite of the Maquine group rangle. It therefore seems questionable to correlate the and the Oambotas Quartzite dips 40° to about 85° N. thick sequence of massive quartzites that form the Serra Also, along the western part of the north slopes of the das Cambotas with the very thin or missing basal Serra do Caraga there is a strong angular discordance quartzite of the Minas Series in this area. between the quartzites of the Maquine and Cambotas Correlation of the quartzite of the Serra das Cam­ Groups. Here the Cambotas Quartzite strikes N. 10° botas with the post-Minas Itacolomi Series is also doubt­ E. and dips east, and the contact strikes west and dips ful, because the quartzite of the Serras Cambotas and north. The contact and the quartzite of the Maquine Geral underlies a sequence of schist, iron-formation, Group in this area are covered by thick jungle growth; and phyllite that is overlain by Caue Itabirite. This however, the contact is exposed on and near the road sequence has been traced westward along the north that leads southward to the Colegio Caraga at N. 6,800, limb of the Gandarela syncline into rocks of unques­ E. 11,500 (pi. 3). Here the quartzites above and below tioned pre-Minas age in the central part of the Quad- the contact are separated by about 0.5-1.5 meters of rilatero Ferrifero. The belt of quartzite in the Serra breccia consisting of angular and subrounded quartzite das Cambotas was therefore correlated with the Eio in a sheared quartz-mica matrix. The contact zone and das Velhas Series (Dorr and others, 1957). However, the quartzites above and below the contact strike west the stratigraphic position of quartzite was not known, and dip 60°-YO° N. except that the quartzite was younger than the schist of the Nova Lima Group and older than the rocks of the TAMANDUi GROUP (UNDIVIDED) Minas Series (Dorr and others, 1957). CAMBOTAS QUARTZITE Simmons and Maxwell (1961) recently studied the The Cambotas Quartzite crops out as a belt of mas­ quartzite and overlying rocks exposed in the Serras sive quartzite that forms the Serra das Cambotas, the Cambotas, Tamandua, and Caraga and presented evi­ southernmost part of the Serra Espinhago, which ex­ dence for a new group of rocks in the upper part of the tends from the eastern part of the Gongo Soco quad­ rangle (pi. 2) northward to the State of Bahia. Kio das Velhas Series which they called the Tamandu& The age of the quartzite that forms the Serras Espin­ Group for the type locality in the Serra do Tamandua hago and Cambotas has been discussed in many reports, in the western part of the. Santa Barbara quadrangle. and evidence has been presented for several different The type section for the Tamandua Group was meas­ ages. Harder and Chamberlin (1915, p. 354) corre­ ured in the eastern part of the Santa Barbara quad­ lated these quartzites with the Caraga Quartzite of their rangle and in the northeastern part of the Gongo Soco Minas Series (table 1), which included some of the quadrangle, at the south end of the Serra das Cambotas quartzites of the Eio das Velhas Series (Dorr and and the west end of the Serra do Tamandua. The section others, 1957, p. 15-22). Bolau (1952, p. 483-487), in a of Tamandua rocks included all the rocks from the base ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 111 of the overlying Caue Itabirite of the Minas Series to the contact with the granodiorite gneiss of the Caete This report Simmons and Maxwell (1961) Complex. Simmons and Maxwell divided this section of rocks into four units (from oldest to youngest): a thick sequence of quartzite beds with a basal quartz-sericite schist member that was named the Cambotas Quartz­ ite for the Serra das Cambotas, and schist, iron-forma­ Minas Series Minas Series tion, and phyllite above the Cambotas Quartzite and below the Caue Itabirite of the Minas Series, which MAJOR UNCONFORMITY(?) were divided into three unnamed units (table 1). UNCONFORMITY The quartzite that forms the northwestern part of the Serra do Caraga in the southwestern part of the Con- ceigao do Kio Acima was placed in the basal quartzite Nova Lima Group Upper Tamandua of Harder and Chamberlin's (1915, p. 351-356) Minas (Dorr and others, 1957) (Simmons and Maxwell, 1961) Series and named the Caraga Quartzite for its type lo­ cality, the Serra do Caraca. Guimaraes (1931, p. 9; 1958, p. 208) correlated the quartzite of the Serra do Caraga with the Itacolomi Series. During the course of MAJOR FAULT MINOR FAULT the present study the quartzite of the northwestern part of the Serra do Caraga was at first, with some reserva­ tions, tentatively correlated with the basal quartzite of the Moeda Formation (Wallace, 1958) of the Minas Cambotas Quartzite Cambotas Quartzite, Series. However, as the result of the study by Simmons (Simmons and Maxwell, 1961) Lower Tamandua and Maxwell (1961), the quartzite of the Serras Cam­ botas and Caraga were correlated with the basal Cambo­ tas Quartzite of their Tamandua Group (fig. 2). The basal quartz-sericite schist unit below the Cambotas UNCONFORMITY UNCONFORMITY Quartzite and the overlying schist, iron-formation, and phyllite have been mapped in this area as part of the Nova Lima Group. This belt of rocks can be traced westward from the south end of the Serra das Cambo­ Nova Lima Group Nova Lima Group tas across the Gongo Soco and Caete quadrangles and (Dorr and others, 1957) (Dorr and others, 1957) into the Nova Lima quadrangle, the type section for Nova Lima (Dorr and others, 1957, p. 15; Gair, 1962, p. 8), where broad areas of the central part of the Qua- drilatero Ferrifero are underlain by rocks of unques­ tioned Nova Lima age. FIGURE 2. Correlation of the TamanduS, Group of Simmons and Dorr (written commun., 1964) suggested that the Maxwell (1961) with the Nova Lima Group of this report. Minas Series as defined by Dorr, Gair, Pomerene, and Eynearson (1957, p. 22-29) and by the joint DNPM- kilometers because of an abrupt stratigraphic wedge USGS team in 1958 (p. 114) should be revised to include out. He stated that the upper contact of this strati- the rocks of the Tamandua Group (Simmons and Max­ graphic wedge out is truncated by segments of two well, 1961) in the eastern part of the Quadrilatero different faults that extend from near N. 7,950, E. 6,400, Ferrifero. According to Dorr, there is not a significant to N. 7,500, E. 8,800, in the Gongo Soco quadrangle angular discordance along the contact between the basal (pi. 2); the lower contact between the schist of the Nova formations of the Minas Series and the upper units of Lima Group and the overlying Cambotas Quartzite is Simmons and Maxwell's Tamandua Group along the cut off by the same faults at N. 7,950, E. 6,400 (pi. 2). eastern part of the Gandarela syncline. He also stated In the eastern part of the Gongo Soco quadrangle (pi. that about 700 meters of Cambotas Quartzite and about 2), the writer mapped a moderate angular discordance 300 meters of the overlying schist, iron-formation, and between the Caue Itabirite of the Minas Series and an phyllite in the upper part of the Tamandua Group at underlying layer of iron-formation of the pre-Minas the south end of the Serra das Cambotas thins west­ Nova Lima Group along the north limb of the Gan­ ward to a knife edge over a horizontal distance of 4 darela syncline at the south end of the Serra das Cam- 112 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL botas. Simmons and Maxwell (1961) correlated this about 0.3 to 1 mm across. Some of the matrix quartz iron-formation with the upper part of their Tamandua fragments show strain patterns in thin sections, and Group. Dorr's rearrangement of the faults along the mortar textures are common. south slopes of the Serra Geral and his truncation of A thick sequence of coarse-grained to very coarse the Cambotas Quartzite along what he considers to be grained white to light-gray sericitic orthoquartzite an abrupt stratigraphic wedge out tend to prove that overlies the lower, conglomerate beds. The orthoquartz­ the Cambotas Quartzite and the overlying schist, iron- ite is crossbedded and consists of well-rounded frosted formation, and phyllite are pre-Minas in age rather than granules of quartzite and quartz 2-1 mm in diameter in Minas, because from N. 7,800, E. 6,000, to N. 8,000, E. a matrix of subrounded and subangular quartz grains 8,000, in the Gongo Soco quadrangle (pi. 2) Cambotas ranging from about 0.2 mm up to the size of the smaller Quartzite is overlain and underlain by schist of the pre- granules. Sericite comprises about 5-10 percent of the Minas Nova Lima Group. Thus, according to his strati- matrix and occurs in discontinuous compact laminae graphic-wedge-out hypothesis, the Cambotas Quartzite and along the boundaries of quartz grains. A few beds intertongues with or is interbedded with the oldest of medium-gray sericite-quartz schist and sericite phyl­ sedimentary rocks of the Quadrilatero Femfero. (See lite are interlayered with orthoquartzite in the south­ %. 2.) western part of the Conceigao do Rio Acima quad­ rangle. The quartzite beds range in thickness from DISTBIBtmON about 0.3 to 2 meters. Crossbedding is a conspicuous Cambotas Quartzite in the southeastern part of the sedimentary feature in the pebble-conglomerate and Conceigao do Rio Acima quadrangle (pi. 3) forms a orthoquartzite beds in the Serra do Caraga. The cross- series of massive cliffs and hogbacks along the north, bedded layers are generally 30-50 cm thick and consist northwest, and west slopes of the Serra do Caraga. In of a series of beveled laminations with the foreset the northeastern part of the Gongo Soco quadrangle laminae dipping in alternate directions, forming a con­ (pi. 2) and the southeastern part of the Antonio dos spicuous herringbone pattern. Scour-and-fill structures Santos quadrangle (pi. 1) Cambotas Quartzite forms are common, and the widespread crossbeddiiig indicates a north-trending line of massive hogbacks along the rapid deposition and frequent current changes. west side of the Serra das Cambotas. At the south end of In the Serra das Cambotas the Cambotas Quartzite the Serra das Cambotas this belt of quartzite narrows, is very thick bedded to massively bedded sericitic quartz­ swings to the west, and becomes the crest of the east ite and quartzite. Weathered surfaces are light gray and end of the Serra Geral (pi. 2). A narrow belt of north- light reddish gray. Fresh rocks are white, light yellow­ trending hogbacks of Cambotas Quartzite occurs within ish gray, and light greenish gray. Beds of sericitic the main body of the granodiorite gneiss of the Caete quartzite weather reddish brown. Weathered surfaces Complex and extends southward from the north bound­ range from very fine grained friable silty texture to ary of the Antonio das Santos quadrangle to the vicin­ rough surfaces of coarse-grained quartz sand. Most beds ity of N. 2,050, E. 4,750 (pi. 1). Other irregular bodies of consist of angular and subrounded quartz grains 1-2.5 Cambotas Quartzite crop out within the granodiorite in mm across. Sericite is present in varying amounts and the northeastern part of the Antonio dos Santos quad­ occurs as thin laminae that impart a rude schistosity to rangle and in the north-central part of the Gongo Soco the quartzite. The coarse-grained quartzite is cut by ir­ quadrangle. regularly spaced north-trending and east-dipping shear LITHOLOQY zones that are flanked by parallel zones of very fine The Cambotas Quartzite of the Serra do Caraga is grained light-greenish-gray quartzite. These very fine made up of pebble-conglomerate beds overlain by inter- grained quartzite bands are dense on fresh break but bedded very coarse grained orthoquartzite and sericitic weather to fine silt-sized material; they are zones of quartzite containing a few interbeds of sericite-quartz cataclasticly crushed quartzite associated with the shear schist and phyllite. zones. The lower, pebble-conglomerate beds are well exposed in the Corrego do Capivara and along the west and Bedding features in the quartzite of the Serra das northwest flanks of the Serra do Caraga. They char­ Cambotas can be seen in the massive cliffs and hogbacks. acteristically weather to rough surfaces of protruding Large-scale crossbedding, scour and fill, and ripple subrounded and rounded quartzite and' quartz pebbles. marks are common. Bedding planes are shown by color The conglomerate consists of quartz and quartzite gran­ banding, laminae of sericite and minor amounts of ules and pebbles 1-10 mm in diameter in a matrix of micaceous hematite, and variation in grain size within angular and subangular quartz grains ranging from beds. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 113

The larger quartz grains are either individual grains THICKNESS AND CONTACT RELATIONS of quartz or aggregates of subrounded smaller grains The thickness of the Cambotas Quartzite in the that have crushed borders. The grains that appear to be central and northeastern parts of the area ranges from an aggregate of smaller grains may be crushed individ­ a few meters on the crest of the Serra Geral at N. ual grains. The smaller grains of quartz, ranging from 7,950, E. 6,350, in the Gongo Soco quadrangle (pi. 2) about 0.1 mm to about 1 mm in the matrix, are angular to 1,200-1,500 meters along the northern and central and apparently are poorly sorted. However, the poor parts of the Serra das Cambotas (pis. 1,2). The quartz­ sorting may be due to shearing rather than to deposition. ite of the Serra das Cambotas is bounded on the south Sericite forms thin laminae and is randomly distributed by the Cambotas fault and on the east, north, and west along grain boundaries in the angular quartz grains of by granodiorite; in most places neither the top nor the the matrix. Minor amounts of plagioclase metacrysts bottom of the quartzite is exposed. Shear zones par­ occur in the quartzite in irregular zones near the contact alleled by narrow zones of cataclasite in the quartzite of the granodiorite. indicate faulting which could have duplicated beds. The very thick bedded to massively bedded quartzite Thus, any estimates of the total thickness of the Cam­ of the Serra das Cambotas grades southward toward botas Quartzite in the Serras Geral and Cambotas are the south end of the Serra das Cambotas into finer meaningless. A thickness of 1,000-1,200 meters is ex­ grained quartzite at the east end of the Serra Geral. posed in a faulted and incomplete section of the Cam­ Here about 700 meters of quartzite is composed of thick- botas Quartzite in the northwestern part of the Serra do bedded very fine grained to coarse-grained white, light- Caraga (pi. 3). gray, light-yellowish-white, light-greenish-gray, and Along the north slopes of the Serra Geral (pi. 2) light-reddish-brown sericitic and ferruginous quartzite. from N. 8,100, E. 6,750, to N. 8,900, E. 10,600, and along The upper 50-70 meters of quartzite is banded by me­ the southwestern part of the Serra das Cambotas (pi. dium-gray laminae of compact micaceous hematite. Un­ 2) from N. 9,100, E. 10,800, to N. 10,600, E. 11,100, Cam­ derlying the ferruginous quartzites is about 400 meters botas Quartzite overlies schist of the Nova Lima Group of thick-bedded white and light-greenish-gray sericitic with slight to moderate unconformity. Along the south quartzite. The sericite is mostly randomly distributed in slopes of the Serra Geral from N. 7,550, E. 5,575, to N. the quartzite, but in some beds it occurs as thin laminae 9,400, E. 13,100, schists of the Nova Lima Group are in fault contact with the Cambotas Quartzite. The under­ parallel to the bedding planes. Some of the beds are lying beds of quartzite are warped to the west and crossbedded and have scour and fill features. The lower dragged into the fault zone of the Cambotas fault, and beds of quartzite contain about 5-20 percent sericite at least 700 meters of quartzite is cut out of the section. that is concentrated in laminae. These sericitic rocks Along the western, northern, and eastern parts of the are deeply weathered and poorly exposed. Serra das Cambotas, Cambotas Quartzite is cut by In the coarser grained beds, the larger grains of granodiorite, gabbro, and diabase of the Caete Complex. quartz range from about 0.5 to 1.5 mm across; however, In most places the contact between the resistant quartz­ the quartzite in this area is crosscut by irregular zones ite and the easily eroded and deeply weathered igneous of cataclastic crushing, and the original grain size of the rocks is poorly exposed or covered by talus or lateritic quartz is not known. Sericite comprises increasing quan­ soil thickly overgrown by scrubby jungle. Although the tities of the lower quartzite beds, and the rocks are contact is covered in most localities, the granodiorite strongly foliated near the south end of the Serra das apparently is intrusive into the quartzite, which is Cambotas. Along the east end of the Serra Geral most partly altered; the quartz is recrystallized, and meta­ of the upper beds of the quartzite are cut out of the sec­ crysts of feldspar are scattered throughout irregular zones near the contact. The contacts between Cambotas tion by the Cambotas fault. Iron-formation conglom­ Quartzite bodies and the granodiorite within the main erates occur in the lower part of the formation along the body of the Caete Complex are also poorly exposed be­ north slopes of the Serra Geral, and there are excellent cause of lateritic soil cover and an overgrowth of scrubby exposures of the conglomerate in the vicinity of N. 8,100, jungle. The bands and irregular masses of Cambotas E. 6,000, and at N. 9,000, E. 12,000 (pi. 2). They consist Quartzite in the granodiorite gneiss all trend generally of angular plates of well-laminated siliceous iron- northward and dip eastward. They are apparently roof formation that are shingled and alined parallel to the pendants and remnants of either faulted blocks or folded foliation of the sericite-quartz schist. Individual plates segments of the main belt of Cambotas Quartzite that range from a few millimeters to about 1 cm in thickness forms the Serra das Cambotas. The quartzite bodies and are about 1-8 cm across. within the Caete Complex are locally warped into open 114 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL folds. Plutons and dikes of gabbro and diabase cross­ overlies the Gandarela Formation and consists of five cut the quartzite along the eastern and southeastern units: the basal Cercadinho Formation (Pomerene, parts of the Serra das Cambotas. The gabbro and di­ 1958a), consisting of ferruginous quartzite and phyllite; abase dikes are intruded into fault and joint planes, the Fecho do Funil Formation (Simmons, 1958) con­ and the dikes are deeply weathered to dark-red soil sisting of dolomitic phyllite, argillaceous dolomite, and that erodes easily and leaves deep trenches within the phyllite; theTabooes Quartzite (Pomerene, 1958b); the quartzite. Mafic dikes in the Serra do Caraga are in­ Barreiro Formation (Pomerene, 1958c), consisting of truded into high-angle thrust faults. The dikes have phyllite and graphitic phyllite; and the Sabara Forma­ sharp contacts with the quartzite. The quartzite has tion (Gair, 1958), consisting of phyllite, schist, and been altered by minor amounts of disseminated pyrite feldspathic quartzite. along a zone 0.5-2 meters wide and is cut by numerous CARACA GROUP fibrous quartz veins. The basal quartzite of the Minas Series was named MINAS SEKIES the Caraga Quartzite by Harder and Chamberlin (1915, Derby (1906, p. 396) first named the Minas Series p. 357) for its type locality in Serra do Caraga. They and included all the metasedimentary rocks above a included the main body of quartzites that forms the Serra do Caraga, which is now known to be composed "gneiss and schist basement." Harder and Chamberlin's of three ages of quartzite: a thin layer of quartzite of (1915, p. 345) Minas Series included part of the coarse the Moeda Formation, exposed along the south and west clastic rocks of Dorr, Gair, Pomerene, and Rynearson's flanks of the range, and two massive layers of quartzite, (1957) Rio das Velhas Series. Harder and Chamberlin both of pre-Minas age, that form the main mass of (1915) divided the Minas Series into five units: the quartzite in the Serra do Caraga. The basal quartzite Caraga Quartzite, Batata! Schist, Itabira Formation, of Minas age was named the Moeda Formation (Wal­ Piracicaba Formation, and Itacolomi Quartzite. Guima- lace, 1958) for the type section in the Serra da Moeda raes (1931, p. 14) showed a major unconformity at the in the western part of the Quadrilatero Ferrifero, be­ base of the Itacolomi Quartzite, separated the Itacolomi cause the ages of the quartzites in the Serra do Caraga from Harder and Chamberlin's Minas Series, and rede­ and elsewhere in the eastern part of the Quadrilatero fined this sequence of quartzite as the Itacolomi Series. Ferrifero were open to question. Harder and Chamber­ In the western part of the Quadrilatero Ferrifero lin (1915) named the phyllitic and schistose unit above Rynearson, Pomerene, and Dorr (1954) described a their Caraga Quartzite the Batatal Schist for the Serra major unconformity at the base of a widespread layer do Batatal. Maxwell (1958) renamed the unit the of quartzite and above a thick sequence of schist and Batatal Formation because it is composed dominantly phyllite. Dorr, Gair, Pomerene, and Rynearson (1957) of phyllite over broad areas of the Quadrilatero and geologists of a joint DNPM-USGS team in 1958 Ferrifero. further restricted the definition of the Minas Series to include the metasedimentary rocks above the uncon­ MOEDA FORMATION formity described by Rynearson, Pomerene, and Dorr DISTRIBUTION (1954) and below the major unconformity at the base Quartzite, conglomerate, and sericite-quartz schist of of the Itacolomi Series (Guimaraes, 1931). The joint the Moeda Formation crop out on the western part of DNPM-USGS team separated the restricted Minas the north limb of the Gandarela syncline. Along the Series into a lower Caraca Group, a middle Itabira southeast limb of the syncline, the Moeda Formation Group, and an upper Piracicaba Group. The Caraga has been mapped from the west boundary of the Con- Group is composed of the Moeda Formation (Wallace, eeigao do Rio Acima quadrangle northwestward to the 1958), consisting dominantly of quartzite and minor east boundary of the Gongo Soco quadrangle. Moeda amounts of phyllite, and the overlying Batatal Forma­ Formation is locally missing in the vicinity of N. 9000, tion (Maxwell, 1958), consisting of phyllite. The Itabira E. 4000 (pi. 3), in the northwestern part of the Con- Group consists dominantly of chemical sediments and eeigao do Rio Acima quadrangle, and near the east includes the Caue Itabirite (Dorr, 1958a), consisting of boundary of the Gongo Soco quadrangle (pi. 2). itabirite a metamorphosed oxide-facies iron-forma­ tion and the overlying Gandarela Formation (Dorr, LITHOLOGY 1958b), consisting dominantly of dolomite interlayered The Moeda Formation is composed of quartzite, con­ with subsidiary iron-formation and phyllite. The con­ glomerate, and quartz-sericite schist. The thicker units tact of the Caue Itabirite and the Gandarela Formation of the formation consist of interlayered conglomerate is gradational. The Piracicaba Group disconforniably and clean white quartzite, and the thinner parts of the ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 115 formation consist of quartz-sericite schist interlayered grains that are about 0.1-0.4 mm across. Many of the with lenses of pebble conglomerate. grains are strained and show mortar textures. Sericite In the western part of the Conceigao do Rio Acima makes up about 5-10 percent of the quartzite and is dis­ quadrangle, excellent exposures of the Moeda Forma­ seminated throughout the rock as plates and discontin­ tion crop out as a series of hogbacks along the south­ uous laminae oriented parallel to the bedding. The peb­ east slopes of the Serra da Pedra Formosa. In this area ble conglomerate consists of well-rounded quartzite and the formation ranges from about 100 to 200 meters in clear, smoky, and milky vein quartz in a matrix of fine­ thickness. The lower and intermediate parts of the for­ grained sericitic quartzite. mation are composed dominantly of white to light-gray Along the western part of the Gandarela syncline very coarse grained thick-bedded quartzite interlayered and locally along the southeastern limb of the syncline, with lenses of pebble and cobble conglomerate. The the Moeda Formation ranges in thickness from a knife very coarse grained quartzite consists of angular to sub- edge to about 25 meters. In these localities the formation angular interlocking quartz grains ranging from about consists of sheared, reddish-gray very fine grained 0.1 to 4.0 mm across; the median size is about 1.5 mm. quartz-sericite schist interlayered with a few lenses of Sericite in minor amounts is distributed throughout the quartzite, and pebble conglomerate. The sericite is con­ quartzite as individual plates and clusters of plates. It centrated in laminae and plates parallel to the schistos- is randomly oriented except in the lower 10-15 meters ity. Pebbles in the conglomerate lenses are sheared and of the formation, where it is alined parallel to the basal flattened, and many of the quartz grains of the quartzite contact of the formation. are crushed and shattered and have cataclastic textures. The interlayered conglomerate beds are lenticular THICKNESS AND CONTACT BELATXONS and do not persist laterally for more than a few hun­ dred meters. They are as much as 4 meters thick and Along the north limb of the Gandarela syncline, the are composed of well-rounded pebbles and cobbles of Moeda Formation ranges from. 20 to 30 meters in thick­ quartzite; clear, smoky, and milky vein quartz; and a ness near the west boundary of the Gongo Soco quad­ few angular fragments of quartz-chlorite schist, all en­ rangle. It thins eastward along a discordant contact closed in a coarse-grained to very coarse grained quartz­ that separates it from the underlying Nova Lima ite matrix. The pebbles and cobbles range from about Group, >and it is cut out of the section in the vicinity of 2 to 250 mm in diameter and have a median size of N. 5,200, E. 4,750 (pi. 2). Along the southeast limb of about 15 mm; the angular to subangular quartz grains the Gandarela syncline at the west boundary of the of the matrix range from about 0.1 mm up to the size mapped area, the Moeda Formation attains its maxi­ of the smaller pebbles. mum thickness of about 200 meters. Northeastward from The upper part of the formation in this area is com­ the Corrego da Mutuca it thins along its lower contact posed of white very coarse grained thick-bedded quartz­ and is either very thin or absent in the northwestern ite consisting of subangular to subrounded quartz part of the Conceic.ao do Rio Acima quadrangle (pi. grains ranging from about 0.2 to about 3.0 mm across; 3). From the south boundary of the Gongo Soco quad­ the median size is about 1.5 mm. rangle (pi. 2) northeastward along the southeast limb Northeastward along the southeast limb of the Gan- of the Gandarela syncline, the Moeda Formation thick­ darela syncline, the conglomerate and very coarse ens from about 10 meters to 100 meters at N. 3,750, E. grained quartzite grade into quartz-sericite schist in the 10,450 (pi. 2). Northeastward from this locality it thins northern part of the Conceigao do Rio Acima quadran­ and is either very thin or missing near the east boundary gle and the southern part of the Gongo Soco quadrangle. of the Gongo Soco quadrangle. In the southeastern part of the Gongo Soco quad­ The upper contact of the Moeda Formation is grada- rangle about 80-100 meters of Moeda Formation is well tional over a zone 1-2 meters thick. The formation exposed along the southeast slope of the Serra da Pedra grades from sericitic quartz schist through sandy Formosa. Here the formation is composed of a basal medium-gray phyllite into dark-gray phyllite and pebble conglomerate 1-3 meters thick overlain by 80-95 graphitic phyllite of the Batatal Formation in most meters of sericitic quartzite interlayered with a few places where the overlying Batatal Formation is 20-30 lenses of pebble conglomerate lithologically similar to meters thick. Near the Corrego da Gandarela (N. the basal conglomerate. The quartzite is weathered in 5,700, E. 1,500, pi. 3), the contact between the Moeda most places to a partly coherent reddish-gray mass of Formation and the overlying argillite of the Batatal quartz granules and sericite plates. Fresh exposures of Formation is a 5- to 10-cm-thick zone of edgewise brec­ the sericitic quartzite are light-gray fine-grained poorly cia consisting of plates of argillite in a sericitic quartz foliated rocks consisting of a mosaic of angular quartz matrix. 116 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL BATATAL FORMATION southeast limb it ranges from as much as 200 meters

DISTBIBUTION near the west boundary of the Conceigao do Bio Acima quadrangle to a knife edge near the east boundary of the The Batatal Formation is covered by soil, canga, Gongo Soco quadrangle. It is apparently cut out of the and talus and is poorly exposed throughout the mapped section by a discordant contact, probably a fault. area, but it can be mapped along an erosional depression The contact between the Batatal Formation and the between outcrops of Caue Itabirite and Moeda For­ overlying Caue Itabirite is gradational through an in­ mation in most places. Batatal Formation occurs along terval of about 0.5-1.5 meters. The uppermost bed of both limbs of the Gandarela syncline. Along the south­ dark-gray phyllite and banded argillite grades upward east limb of the syncline it has been mapped for more through a poorly laminated ferruginous and quartzitic than 18 kilometers, from the west boundary of the phyllite into well-laminated Caue Itabirite. mapped area to N. 5,000, E. 12,000, near the east bound­ ary (pi. 2). Northeastward from this locality to the ITABIRA GROUP east boundary of the Gongo S6co quadrangle, the CAUE ITABIRITE Batatal Formation is missing from the section. It is DISTEIBUTION apparently cut out of the section by a fault zone Caue Itabirite is resistant to erosion and forms the between Caue Itabirite and the underlying Moeda For­ higher ridges of the area; it crops out along both limbs mation. Near the east boundary of the Gongo Soco of the Gandarela syncline. Along the southeast limb of quadrangle, lenses of phyllite occur over 'and under the syncline it forms the northeast-trending crest of the Moeda Formation, and they are mapped as Caraga Serra da Pedra Formosa, and it has 'been mapped for Group (undivided). Along the north limb of the Gan- more than 18 kilometers from the west boundary of the darela syncline, the Batatal Formation has been mapped Conceigao do Eio Acima quadrangle to the east bound­ from the west boundary of the Gongo Soco quadrangle ary of the Gongo Soco quadrangle. Along the north to the vicinity of Camara gold mine, where it is cut limb of the Gandarela syncline the itabirite forms an out of the section along a discordant contact that sepa­ east-northeast-trending high ridge along the south rates Caue Itabrite from the underlying rocks of the slopes at the Serra Geral, and it has been mapped across Nova Lima Group. the central part of the Gongo Soco quadrangle. LITHOLOGY The name "itabirite" was apparently used very The Batatal Formation is composed dominantly of early to describe a "whetstone," and Eschwege (1822) phyllite, graphitic phyllite, argillite, and a few lenses of adapted it to describe high-grade hematite ore (Guild, recrystallized chert and ferruginous chert. The thicker 1957, p. 14). Since 1822 it has been used to describe a parts of Batatal Formation occur in the western part wide variety of iron-rich rocks ranging from well- of the Conceic,ao do Eio Acima quadrangle and are com­ laminated iron-formation to high-grade hard hematite posed dominantly of banded light- and medium-gray ore. It is currently accepted as the name for metamor­ argillite containing numerous cubic pseudomorphs of phosed oxide-facies iron-formation that consists of al­ specular hematite after pyrite. In the upper part of the ternating laminae of recrystallized iron oxide, chert, section the argillite is interlayered with lenses of recrys­ and carbonate containing 25-66 percent iron (Dorr and tallized chert and ferruginous chert. Quartz crystals Barbosa, 1963, p. 18). Laminated rocks containing less as much as 2 cm across and specular hematite crystals than 25 percent iron are considered by Dorr and Bar­ up to 5 mm across occur in the lenses. The argillite in bosa to be ferruginous quartzite, and rocks containing most places has weathered to saprolite composed of more than 66 percent iron are classified as high-grade banded light- and medium-gray clay. hematite ore. The thinner parts of the formation in the mapped Caue Itabirite may be generally classified as either area are composed of foliated and intensely sheared siliceous itabirite or dolomitic itabirite, depending upon dark-gray phyllite and graphitic phyllite containing the composition of the light-colored laminae. Both types a few ferruginous phyllite lenses. of itabirite occur in the area mapped, and they are inter- laminated in the lower and middle parts of the forma­ THICKNESS AND CONTACT RELATIONS tion. In most exposures both types of itabirite are The Batatal Formation ranges in thickness from deeply weathered to a partly coherent banded friable about 30 meters to a knife edge along the western part mass of granular quartz, hematite, and powdery limo- of the north limb of the Gandarela syncline. Along the nite. Hard unweathered poorly laminated iron-poor southeast limb of the syncline the Batatal Formation siliceous itabirite is widely exposed in a series of jagged is variable in thickness and is locally absent. Along the hogbacks extending northeastward from near the Piao ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO EIO ACIMA QUADRANGLES 117 prospect, N. 5,100, to N. 6,100, E. 850, in the western tabular plates and irregular grains of hematite inter- part of the Conceigao do Eio Acima quadrangle (pi. 3). crystallized with minor amounts of quartz. The flakes This iron-poor itabirite occurs near the hard hematite and granules range from 0.02 to 0.2 mm across, and the ore body of the Piao prospect and apparently has been quartz ranges from 0.01 to about 0.3 mm across. The hydrothermally altered, as it is more resistant to sur­ light-colored laminae consist of interlocking quartz face weathering than the adjacent dolomitic itabirite. grains 0.02-0.2 mm across and a few grains of hematite About 300 meters northeast of the hogbacks of hard enclosed in individual quartz grains or cutting across siliceous itabirite, about 300 meters of hard poorly the boundaries of quartz grains. Magnetite octahedra laminated iron-poor dolomitic itabirite crops out along were not observed either in the outcrops or in the thin the base of a steep canyon cut into the Serra da Pedra sections of the siliceous itabirite. Formosa by the Corrego da Gandarela (N. 6,200, E. Excellent exposures of unweathered or only slightly 1,100, pi. 3). weathered dolomitic itabirite occur in the Corrego da The Caue Itabirite is well exposed along the south Gandarela near N. 6,700, E. 1,200 (pi. 3). Thin sections slopes of the Serra Geral in the vicinity of the Gongo from a 300-meter-thick unit of dolomitic itabirite in Soco iron mine. It consists mostly of dolomitic itabirite the upper part of the formation show that the rock con­ in the upper part of the section and interlaminated sili­ sists of iron-rich laminae made up of plates and irregu­ ceous and dolomitic itabirite in the middle and lower lar grains of hematite, and magnetite octahedra inter- parts of the section. Northeastward from the Gongo crystallized with quartz and dolomitic carbonate. The Soco iron mine to the east boundary of the quadrangle, hematite and magnetite grains range from 0.02 to 0.4 the lower part of the Caue Itabirite is weathered sili­ mm across and have a median size about 0.2 mm. Mag­ ceous magnetic itabirite that grades upward into inter- netite octahedra are enclosed by or cut across grain laminated siliceous and dolomitic itabirite that in turn boundaries of quartz, dolomite, and hematite. The light- grades into dolomitic itabirite in the upper part of gray laminae consist of a mosaic of quartz grains en­ the formation. closing grains and crystals of dolomitic carbonate; West of the Gongo Soco iron mine to near the Gongo plates of sericite, biotite, and hematite; and a few crys­ Soco gold mine, the Caue Itabirite is cut by numerous tals of magnetite. Quartz grains range from 0.02 to irregular bodies of hydrothermally enriched itabirite about 0.2 mm across. Dolomitic carbonate grains are ir­ (hard and soft hematite ores). Zones of both hard and regular or rhombohedral and range from 0.01 to about soft hematite ores also occur along the crest of the 0.2 mm across. Dolomitic carbonate comprises as much Serra da Pedra Formosa from near the Piao prospect as 40 percent by volume of the light-colored laminae (N. 4,700, E. 200, pi. 3) to near the Santo Antonio mine and from 5 to 20 percent by volume of the itabirite. (N. 6,350, E. 1,300). The enriched are dis­ Sericite makes up about 0.5-3.0 percent of itabirite and cussed on page 136. forms prismatic plates that cut across the grain bound­ Hard dolomitic itabirite is also well exposed along a aries of quartz and dolomitic carbonate. Light-green small corrego that leads south from the Gongo Soco iron biotite occurs in a few thin sections as prismatic grains mine in the Gongo Soco quadrangle. Elsewhere in the and as clusters of crystals and is an accessory mineral. quadrangle the Caue Itabirite either is weathered sili­ The dolomitic itabirite is coarsely laminated (5-20 mm ceous itabirite or is covered by lateritic soil and canga. thick), and iron-rich bands and light-colored bands Itabirite is a laminated dark- and light-gray rock. are crudely developed. Dark-gray iron-rich laminae made up of flakes and Although hard dolomitic itabirite is exposed in granules of hematite and minor amounts of magnetite deeply eroded rocks and hard siliceous itabirite crops alternate with light-gray laminae consisting of a out in the vicinity of hard hematite ore bodies, the mosaic of quartz and (or) dolomite grains. Segregation Caue Itabirite in most areas is weathered to a friable of the minerals in the laminations is not complete; mass of banded hematite and quartz that crumbles with some of the dolomite and quartz grains occur in the the slightest blow. hematite-rich laminae, and a few flakes and granules of In a few localities itabirite weathers to a mass of hard hematite and magnetite octahedra are scattered through plates of iron-oxide and disaggregated granular quartz. the quartz and dolomite laminae. Laminations range This type of weathered itabirite is locally called from about 1 to 20 mm in thickness. Individual laminae "Chapinha," and it is used in the local smelters because maintain a constant thickness for several meters in it is easily concentrated by screening out the quartz. many exposures. The most graphic evidence of surface weathering of Under the microscope, the iron-rich laminae in sili­ hard dolomitic itabirite occurs in a steep-walled canyon ceous itabirite are seen to be a mosaic of interlocking cut through the Serra da Pedra Formosa near the west 118 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL boundary of the Conceigao do Rio Acima quadrangle the Gongo Soco quadrangle. Westward from N. 5,050, (N. 6,200, E. 1,100, pi. 3) by the Corrego da Gandarela. E. 4,700, the Caue Itabirite thickens to about 400 meters Here at least 300 meters of hard dolomitic itabirite ex­ near the west boundary of the Gongo Soco quadrangle, posed along the base of the canyon grades upward along and eastward from the same location it thickens to about the canyon walls into partially softened itabirite, which 300 meters near the Gongo Soco iron mine. Eastward in turn grades into disaggregated soft siliceous itab­ from about N. 5,150, E. 4,800 (pi. 2), to the east bound­ irite at the top of the canyon. Thin sections of par­ ary of the Gongo Soco quadrangle, the Cane1 Itabirite tially weathered itabirite at the top of the canyon show overlies schist and ferruginous quartzite of the Nova only quartz in the light-colored laminae and iron oxide Lima Group. Discordant features of this contact sug­ in the dark-colored laminae, whereas thin sections of gest that some of the lower beds of the Caue Itabirite hard itabirite from the same layer at the base of the may be cut out of the section by fault displacement canyon contain about 10 percent by volume of dolo­ along this contact in the eastern part of the north limb mitic carbonate in both the light-colored laminae and of the Gandarela syncline. Eastward from the Gongo the dark-colored iron-oxide laminae. A channel sample Soco iron mine, the Caue Itabirite thins from about 300 across 6 meters of the same band of soft itabirite assayed meters to about 200 meters near the east boundary of 40.2 percent iron and 41.2 percent SiO2. In comparison, the Gongo Soco quadrangle; however, a fault cuts out a chip sample of hard itabirite from the same band at some of the upper beds of the formation. the base of the canyon assayed 29.4 percent iron, 47.8 In the western part of the Conceigao do Rio Acima percent SiO2) 0.5 percent A12O3, 3.2 percent CaO, 1.6 quadrangle, the Caue Itabirite is about 650 meters thick percent MgO, and a trace of manganese. Judging from near the Piao prospect. Northeastward from the Piao the soft itabirites at the top of the canyon that form prospect along the crest of the Serra da Pedra Formosa, the crest of the Serra da Pedra Formosa, and their Caue Itabirite thins to about 300 meters near Pedra downward gradation into hard itabirite, it is apparent Vermelha (N. 750, E. 6,000, pi. 2) in the southern part that surface waters have disaggregated the soft of the Gongo Soco quadrangle. Northeastward from itabirite. The comparative thin-section studies and Pedra Vermelha, Caue Itabirite crops out in a series of analyses of the hard and soft itabirite indicate that fault blocks, and little is known about its thickness. Al­ some of the quartz and most, if not all, of the dolomitic though the Caue Itabirite is locally thinned by faults, carbonate have been leached from dolomitic itabirite its thickness ranges from about 650 meters at the west near the surface; this leaching has resulted in residual boundary of the Conceigao do Rio Acima quadrangle enrichment and disaggregation of the itabirite. to about 300 meters in the central and southern parts of A small part of the iron oxide in the dolomitic ita­ the Gongo Soco quadrangle. It reaches an apparent birite is magnetite, whereas the iron oxide in the soft thickness of 800 meters southwest of the Piao prospect itabirite is dominantly hematite. Both types of ita­ in the Gandarela quadrangle (J. E. O'Rourke, unpub. birite are magnetic, and most of the magnetic soft ita­ data). birites probably are derived from the weathering of The upper contact of the Caue Itabirite is grada- dolomitic itabirite. Exposures of magnetic soft sili­ tional over a 20- to 30-meter-thick zone into dolomite of ceous itabirite unsually contain ocher and talc, which the Gandarela Formation. This gradational contact is suggest weathering of dolomitic itabirite. Pomerene in most places deeply weathered and covered by soil; it (1964, p. 13), however, noted that magnetic siliceous is exposed in only two places. An exceptional view of itabirite is a surface phenomenon. He noted that the this gradational contact zone can be seen in a massive subsurface soft weathered itabirite is rarely magnetic overhanging cliff, locally called Pedra Vermelha, on the and suggested that weathering or possibly reduction by northwest slopes of the Serra da Pedra Formosa at N. sunlight may cause magnetic itabirite. 950, E. 5,800 (pi. 2). Here, well-laminated dolomitic itabirite grades into about 15 meters of coarsely lami­ THICKNESS AND CONTACT RELATIONS nated alternating gray dolomite, specularite, and The apparent stratigraphic thickness of the Caue Ita­ quartz, which in turn grades into about 10 meters of birite within the Conceigao do Rio Acima and Gongo gray quartzitic dolomite stratigraphically overlain by Soco quadrangles ranges from about 100 to 650 meters. massive beds of red and pink dolomite of the Gandarela The Caue Itabirite crops out along both limbs of the Formation. Another excellent exposure of this contact Gandarela syncline. Along the north limb on the south zone between the Caue Itabirite and dolomite of the slopes of the Serra Geral the thickness is variable; it Gandarela Formation crops out in a canyon cut through ranges from about 100 meters in the vicinity of N. 5,050, the Serra da Pedra Formosa at N. 6,250, E. 1,050, in the E. 4,700 (pi. 2), to 400 meters near the west boundary of southwestern part of the Conceicjio do Rio Acima ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 119 quadrangle (pi. 3). Here well-laminated hard dolo- crop out on the southeast limb of the Gandarela syncline mitic itabirite of the Caue Itabirite grades upward into except near N. 6,150, E. 12,300 (pi. 2). Dolomite also a zone of coarsely laminated ferruginous gray dolomite crops out on the north limb of the Gandarela syncline about 25 meters thick, which in turn grades into mas­ and forms a belt approximately 1 kilometer wide that sive beds of varicolored gray and red dolomite of the has been mapped completely across the Gongo Soco Gandarela Formation. quadrangle (pi. 2). Because the gradational zone between the overlying The dolomite of the Gandarela Formation is deeply dolomite of the Gandarela Formation and the under­ weathered to saprolite and lateritic soils and forms a lying laminated itabirite of the Caue Itabirite is in rounded topography of moderate relief. Low cliffs of most places covered by lateritic soils and canga, the con­ resistant varicolored banded dolomites protrude from tact zone has been mapped either where the light-tan the saprolite and lateritic soils in the Rio do Socorro argillaceous soils of the weathered dolomite grade into valley and along the lower slopes of the Serra Geral. reddish-brown granular soils of the weathered dolo- Dolomite is exposed in prominent cliffs, hogbacks, and mitic itabirite or between outcrops of dolomite and ledges in the vicinity of the Corrego da Gandarela well-laminated dolomitic itabirite. Where the contact (N. 6,300, E. 500, pi. 3). Northeastward from the Cor­ is covered by canga, the contact between the Caue rego da Gandarela, the dolomites of the Gandarela Itabirite and the dolomite of the Gandarela Formation Formation underlie an erosional surface overlain by can only be projected across the canga. canga and form the low rolling topography of the La- Dorr (1958 a, b), proposed that the contact between goa Grande plateau (pi. 3). the Caue Itabirite and the Gandarela Formation be drawn where siliceous itabirite becomes subordinate to LITHOLOGY dolomitic itabirite. This definition is useful in areas in The Gandarela Formation is made up of thick-bedded the QuadriMtero Ferrifero where the Caue Itabirite gray, red, pink, purple, white, and cream-colored dolo­ consists mostly of siliceous itabirite. In the mapped mite interbedded with itabirite and phyllite. The lower area, however, dolomitic itabirite comprises a large part strata are medium-gray and white dolomite banded by of the Caue Itabirite, and it is not possible to divide the laminae of quartz and hematite. These strata grade up­ Itabira Group on the basis of this criterion. ward into massive white and cream-colored dolomite GANDARELA FORMATION banded by very light gray to dark-gray mica laminae. The upper strata are dark-red dolomite interbedded The Gandarela Formation in the Conceigao do Rio with laminated purple, pink, and light-gray dolomite. Acima and Gongo Soco quadrangles consists of dolo­ The lower dolomite is banded by numerous laminae mite, dolomitic itabirite, siliceous itabirite, and a few of quartz and specular hematite. Magnetite octahedra interbeds of phyllite and conglomerate. are disseminated in the dolomite layers and in the hema­ In 1954, during the early stages of mapping in the tite laminae. Thin lenses of well-laminated dolomitic Quadrilatero Ferrifero, J. E. O'Rourke (unpub. data) itabirite are interlayered at many levels within the lower mapped the lower dolomite beds of the type locality in dolomite beds. Phyllite layers are concentrated in the the Gandarela quadrangle as a separate unit of what lower part of the formation and are in most places ex­ was then called the Itabira Formation. He was the posed only as weathered soils; however, good exposures first to recognize and map the dolomite as a separate of grayish-green phyllite crop out on a ridge southeast formation and he subdivided the dolomite into three of the Capim Gordura manganese prospect (N. 3,600, members on the basis of color. A similar division in the E. 1,050, pi. 2). quadrangles mapped for this report is not possible be­ The intermediate dolomite beds of the Gandarela For­ cause color changes are not confined to a narrow strati- mation are well exposed in numerous ledges, hogbacks, graphic interval. and quarries along the south slopes of the Serra Geral DISTRIBUTION in the central and eastern parts of Gongo Soco quad­ rangle. The fresh dolomite is light gray to cream col­ Dolomite of the Gandarela Formation crops out on ored, fine to medium grained and thick bedded and is the southeast limb of the Gandarela syncline and has banded by laminae of light-gray to cream-colored phlog- been mapped for more than 17 kilometers. Near Lagoa opite. A few beds of ferruginous dolomitic quartzite das Antas (pi. 2), the dolomite is offset by a northeast- are interbedded with the micaceous dolomite. The mica­ trending fault, but it then continues to the northeast ceous dolomite weathers to a medium-gray granular to where it is cut off by another fault at N. 3,900, E. surface, and the phlogopite laminae weather to talc 1,000 (pi. 2). Northeastward, the dolomite does not and reddish-brown limonite. 120 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL Thin sections of the cream and white dolomite show red dolomite show it to contain 0.07-1.2 percent a mosaic of interlocking dolomite (85-95 percent) and manganese as compared with about 0.2 percent manga­ quartz (5-15 percent) grains ranging from 0.04 to 0.5 nese in the lighter colored dolomite. The manganese mm across. The phlogopite has a light-gray to light- must occur in the opaque mineral because no other man­ cream pearly luster and comprises about 5-10 percent ganese-bearing minerals were identified in the thin- of the rock by volume; it occurs in regular and wavy section studies. compact laminae ranging from 1 to 5 mm in thickness. The finely laminated purple dolomite and the dark- Pyrite is sparsely disseminated in most dolomite beds red dolomites are cut by numerous veinlets of specular and is not associated with fractures or linear structures. hematite ranging from about 0.5 to 5 mm in width. The It occurs as small cubes and aggregates of cubes rang­ veinlets occur along the laminations and in crosscutting ing from a fraction of a millimeter to about 10 mm fractures and joint planes. The specular hematite may across. Quartz veins are common and occur as fillings in be indigenous iron that was mobilized by through-going joints and in minor faults. hydrothermal fluids, or it may have been introduced by Ferruginous and dolomitic quartzite beds consist of iron-bearing hydrothermal fluids. The dolomite contains about 55-65 percent angular quartz intercrystallized sufficient iron to account for a mobilization and transfer with about 35-45 percent angular and rhombohedral of iron within the formation. Evidence for large-scale dolomite. They are crudely laminated by compact specu­ hydrothermal transfer of iron can be seen in the Cabega lar hematite containing some octahedral magnetite. de Fen-o mine, where the dolomite is replaced along the Magnetite is also sparsely disseminated in the dolomitic axial zone of a fold by hard hematite; this relation indi­ quartzite layers. cates an outside hydrothermal source for the iron. How­ Massively bedded very fine grained dark-red man- ever, the specular hematite in the veinlets probably was ganiferous dolomite interbedded with layers of finely derived from the iron contained in the enclosing laminated very fine grained purple and light-gray man- dolomite. ganiferous dolomite comprise the upper part of the A conglomerate bed occurs about 50 meters below the Gandarela Formation in the western part of the area. top of the formation in the face of the Socorro quarry Outcrops of these manganiferous dolomites are rare as at N. 3,600, E. 4,600 (pi. 2). The conglomerate is about 1 the rocks generally are deeply weathered to bluish- meter thick and is composed of white and gray frag­ black plastic manganiferous lateritic soil and nonplas- ments of dolomite 2-20 cm across in a dark-gray poorly tic black manganiferous material a few meters below foliated matrix consisting of dolomite, iron oxide, and the surface. Excellent exposures of the dark-red dolo­ mica. Another layer of conglomerate is exposed in the mite and laminated purple and light-gray dolomite can face of a quarry south of the Pedra Vermelha at N. 1,000, be seen in the Socorro quarry at N. 3,600, E. 4,600, and E. 5,950, in the Gongo Soco quadrangle (pi. 2). This in the Paracatu quarry at N. 4,110, E. 4,650, in the cen­ conglomerate bed is about 50 cm thick and is composed tral part of the Gongo Soco quadrangle (pi. 2). Thin of rounded and subrounded gray and red dolomite frag­ sections of the dark-red dolomite show the rock to be ments 5-15 cm across in a dark-gray phyllite matrix. an interlocking mesh of dolomite grains, 0.01-0.04 mm The dolomite fragments in both conglomerate layers are across, enclosing and replacing widely scattered quartz stretched and compressed, and the long axes lie in the grains that make up about 1-3 percent of the rock by plane parallel to the foliation in the matrix. The con­ volume. Disseminated throughout the mosaic of dolo­ glomerates exposed in these two quarries may represent mite and quartz are minute ragged flakes of an unidenti­ the same horizon, but limited exposures make it difficult fied opaque mineral ranging from 0.004 to 0.02 mm to use them as a marker bed. Similar conglomerates in across. Under the microscope the banded purple and the Gandarela Formation have been found elsewhere in light-gray dolomite is seen to be composed of alternat­ the Quadrilatero Ferrif ero. ing light-gray laminae of dolomite grains 0.01-0.06 mm A thick layer of manganiferous siliceous itabirite across and purple laminae consisting of a mosaic of crops out in the upper part of the formation in a low about 90 percent dolomite, 5 percent quartz, and widely hogback at N. 5,060, E. 6,750 (pi. 2). The itabirite con­ disseminated anhedral flakes of an unidentified opaque sists of dark-brownish-black laminae made up of gran­ mineral. The dolomite and quartz range from 0.01 to ules and flakes of hematite and magnetite and minor about 0.04 mm across, and the opaque mineral is about amounts of manganese oxide. The light-gray laminae 0.004-0.01 mm across. Although this mineral has not consist of granular quartz and a few scattered flakes and been identified, it is apparently a manganiferous min­ granules of hematite and magnetite. The hematite and eral that imparts a darker color to the dark-red and pur­ magnetite grains range from 0.02 to 0.10 mm across, ple laminations of the dolomite. Analyses of the dark- and the quartz grains range from 0.01 to about 0.03 mm ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 121

across. The dark laminae contain fine black powdery The light-tan honeycombed aluminous soils occur at manganese oxide that smudges the hand. Analyses of many places throughout the Lagoa Grande Plateau. The the itabirite show it to contain 3.2 percent manganese. residual soil contains as much as 35 percent A12O3 and is Dolomite of the Gandarela Formation weathers to derived from weathered dolomite. No subsurface expo­ lateritic soils of two main types: light-tan aluminous sures of fresh dolomite were observed, and the nature soil and manganiferous soil that ranges from reddish of the weathering process and the type of dolomite that brown to bluish black and forms a plastic, wadlike weathers to these highly aluminous1 soils is not known. material near the surface. A few meters below the It is known only that these aluminous soils occur near surface, the manganiferous plastic soil grades abruptly the eroded and undercut perimeter of the canga fields into a black vitreous nonplastic material that resembles where there is rapid downward percolation of meteoric low-grade lignite in color and texture; however, when waters. Thus, the 'alumina-rich residual soil may be struck with a hammer it splatters in the same manner the residue of micaceous dolomite leached of its more as mud. This material was first recognized as a weather­ soluble constituents such as CaO and MgO, leaving the ing product of dolomite by Guild, who named it "splash less soluble A12O3. The mica probably accounts for the rock" (Guild, 1957, p. 42). Outcrops of "splash rock" A12O3. are initially dry to the touch, but when a fragment is THICKNESS AND CONTACT EELATIONS squeezed between the fingers it collapses to a mass of A thickness of about 70-1,000 meters is indicated for black mud. This material is formed in place by meteoric the dolomite of the Gandarela Formation in the Gongo waters leaching the carbonate from the fine-grained Soco and Conceigao do Rio Acima quadrangles. The for­ manganiferous dolomite. The iron oxide probably is in mation attains its maximum thickness on the south­ part indigenous iron, and an undetermined amount may east limb of the Gandarela syncline near the west bound­ have been introduced by meteoric water. ary of the Conceigao do Eio Acima quadrangle (pi. 3). Excellent exposures of "splash rock" can be seen Northeastward from this locality it thins to about 750 within a dolomite layer in the Socorro quarry (N. meters in the vicinity of Lagoa Grande (N. 13,300, E. 3,650, E. 4,650, pi. 2). Here the black "splash rock" 4,600, pi. 3.) Northeastward from Lagoa Grande to the is exposed as lenses overlain and underlain by fresh east boundary of the Gongo Soco quadrangle the Gan­ dolomite. The lenses of "splash rock" in the Socorro darela Formation apparently thins to about 100 meters; quarry contain relict lenses and rounded nodules of however, faulting makes any estimate of its true strati- partially weathered dolomite. The boundaries between graphic thickness meaningless in the eastern part of the the "splash rock" and the dolomite are narrow grada- mapped area. tional zones of brown flaky material 1-2 cm wide. The Along the north limb of the Gandarela syncline the dolomite appears to have weathered in place, without dolomite ranges from about 100 to 800 meters in thick­ slumping, leaving a microscopic boxwork of iron and ness. It is thickest near the west boundary of the Gongo manganese oxides of the same volume as the unweath- Soco quadrangle and thins northeastward to about 500 ered dolomite. The weathered products ,when saturated meters near the village of Gongo Soco (N. 5,400, E. with water have a specific gravity probably half that 8,000, pi. 2). Eastward from Gongo Soco the dolomite of the unaltered dolomite. Analyses of fresh and altered is faulted and folded and its stratigraphic thickness is dolomite from which the "splash rock" is derived are not known. Simmons reported (written commun.) a shown in table 2. thickness of 30-70 meters for the Gandarela Formation TABLE 2. Analyses (in percent) of fresh dolomite and weathered along the northeasetern part of the Gandarela syncline dolomite from the Socorro quarry (N. S,6BO, E. 4,650, pi. #) Gongo 86co quadrangle in the Santa Barbara quadrangle east of the Gongo [Analyst: Dr. Cassio Pinto, Departmento Nacional da ProducSo Mineral, Belo Soco quadrangle. Horizonte] Like the Caue Itabirite, tho Gandarela Formation is la 3a thickest at the west boundary of the area along both limbs of the Gandarela syncline and thins northeast­ CaO_. __ 30.3 9.3 30.4 5.4 MgO. 20.8 8.1 20.7 8.1 20.1 4.3 ward across the area along both limbs of the Gandarela MnO2 1.4 37.8 1.6 35.1 1.2 29.5 FeO_. . 1.1 17.8 1.0 16. 8 3. 6 40.5 syncline. The proportionate thinning of both forma­ SiO2- ____.______.__ .2 3.5 .2 4.4 .2 1.7 A1203- .-__.______.3 .5 .1 .5 .4 .5 tions of the Itabira Group along the northwestern part Loss on ignition- __.______46.5 25.0 46.4 30.0 46.0 18.4 of the Gandarela syncline may indicate tectonic thin­ Total ______100.6 102.0 100.4 104.2 101.1 100.3 ning by plastic flow of rock in the limbs of the Gan­ 1. Fresh dark-red dolomite. la. Black "splash rock" derived from red dolomite of sample 1. darela syncline. A second possibility could be deposi- 2. Fresh dark-red dolomite enclosing lens of black "splash rock." tional thinning of both formations northeastward 2a. Black "splash rock" lens enclosed by dark-red dolomite of sample 2. 3. Lens of dark-red dolomite enclosed by black "splash rock." 3a. Black "splash rock" enclosing dark-red dolomite of sample 3. along the syncline; however, it is unlikely that both 122 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL formations would show almost identical depositional north limb of the Gandarela syncline as hogbacks and thinning. A third possibility is that the upper beds of ledges and as bluffs of quartzite and ferruginous quartz­ the Gandarela Formation were eroded prior to the ite. In the southeastern part of the Gongo Soco quad­ deposition of the Cercadinho Formation. At the So- rangle and the northwestern part of the Conceic.ao do corro dolomite quarry (N. 3,600, E. 4,600, pi. 2) and near Rio Acima quadrangle on the southeast limb of the the west boundary of the Gongo Soco quadrangle (N. Gandarela syncline the quartzite and ferruginous 2,000, E. 100, pi. 2) dark-red dolomite beds of the Gan­ quartzite of the Cercadinho Formation are deeply darela Formation are overlain by ferruginous quartzite weathered and poorly exposed. At the east boundary of the Cercadinho Formation. In contrast, the upper of the Gongo Soco quadrangle the Cercadinho Forma­ 'beds of the Gandarela Formation at a quarry at N. tion forms the axial zone of the Gandarela syncline.

3,650, E. 8,600 (pi. 2), are white and cream-colored LITHOLOGY dolomite that is stratigraphically overlain by ferrugi­ The Cercadinho Formation is composed of ferrugi­ nous quartzite of the Cercadinho Formation. Also, from nous quartzite, quartzite, phyllite, and quartz-sericite- near N. 4,850, E. 8,100 (pi. 2), to the east boundary of chlorite schist. Individual beds lens out and can be the Gongo Soco quadrangle (N. 7,700, E. 13,050, pi. 2) traced for only a few hundred meters. In the western quartzite and quartz-sericite schist of the Cercadinho part of the Gongo Soco and Conceicao do Rio Acima Formation overlie white, light-gray, and cream-colored quadrangles, the formation consists principally of me­ dolomite that is lithologically similar to the dolomite dium to coarse-grained quartzite and ferruginous beds that form the intermediate part of the Gandarela quartzite and interbeds of sericite phyllite and reddish- Formation in the western part of the Gongo Soco and brown chlorite phyllite. Eastward along both limbs of Conceigao do Rio Acima quadrangles. the Gandarela syncline numerous lenses and interbeds PIRACICABA GROUP of phyllite and ferruginous phyllite occur in the section. The Piracicaba Group comprises the upper part of In the western part of the area the upper beds are the Minas Series and is overlain unconformably by the white, light-yellow, buff, and light-gray medium- to Itacolomi Series (Guimaraes, 1931, p. 8) in other areas coarse-grained thick-bedded quartzite containing minor of the Quadrilatero Ferrif ero. Harder and Chamberlin laminae of sericite, chlorite, and hematite. The quartz­ (1915, p. 362) named the Piracicaba Formation and in­ ite beds are separated by thin partings of dark-gray cluded part of the underlying Gandarela Formation in graphitic phyllite and sericite phyllite. A unique sand­ it. Dorr, Gair, Pomerene, and Rynearson (1957) ele­ stone bed occurs near the top of the formation at N. vated this formation to group status and included the 3,900, E. 6,150 (pi. 2). It consists of rice-shaped frosted clastic rocks above the chemical sedimentary rocks of granules of quartz whose long axes are crudely alined the Itabira Group and below the quartzite of the Itaco­ with the bedding planes. This type of quartzite has been lomi Series. found only in the Cercadinho Formation in the Quadri­ The Piracicaba Group consists of quartzite, phyllite, latero Ferrifero (Pomerene, 1964, p. 23). ferruginous quartzite, and feldspathic quartzite in the The lower beds of the Cercadinho Formation in the Conceigao do Rio Acima and Gongo Soco quadrangles. western part of the area consist of poorly laminated In the western part of the Quadrilatero Ferrifero the medium-gray and purplish-gray ferruginous quartzite Piracicaba Group has been divided into five formations, separated by 1- to 50-cm-thick interbeds of chlorite but in these quadrangles only two of these formations phyllite and f erruginous sericite phyllite. The quartzite have been recognized. The basal formation is composed weathers to reddish-brown friable soil consisting of li- of quartzite, ferruginous quartzite, and phyllite and is monite, mica, and quartz granules. The chlorite phyllite correlated with the Cercadinho Formation described by partings in many places show finely crenulated laminae. Pomerene (1964, p. 20-25). The upper formation is Good exposures of the lower part of the section can be composed of phyllite, feldspathic quartzite, and quartz- seen in a roadcut near N. 3,900, E. 6,000 (pi. 2). sericite schist. Its stratigraphic position and lithology The amount of argillaceous material in the Cerca­ are similar to those of the Sabara Formation described dinho Formation increases northeastward, and in the by Gair (1958), and it is therefore correlated with the eastern part of the Gongo Soco quadrangle the forma­ Sabara Formation. tion is made up of well-laminated ferruginous quartzite interbedded with sericite phyllite and ferruginous CERCADINHO FORMATION quartz-sericite schist. DISTRIBUTION The ferruginous quartzite contains widely varying The Cercadinlio Formation is exposed along the amounts of granular and micaceous hematite. The hema­ lower part of the south slopes of the Serra Geral on the tite occurs mostly in thin laminae in the quartzite, but ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 123 in a few beds it is evenly disseminated throughout the the quadrangles the contact is covered with soil, but it quartzite. The hematite laminae range from 1 mm to has been mapped within about a 10- to 20-meter strati- about 5 mm in thickness and locally show low-angle graphic interval on the basis of soils and outcrops. crossbedding. Plates and granules of specularite are ar­ SABARA FORMATION ranged in poorly denned bands and comprise about 5-15 percent of the quartzite by volume. DISTRIBUTION Thin sections of unweathered ferruginous quartzite Phyllite, schist, and feldspathic quartzite of the Sa­ show it to be a subangular mosaic of quartz grains rang­ bara Formation are exposed in the Socorro valley as a ing from 0.05 to 1.0 mm across and a few porphyro- broad belt that extends from the east boundary of the blasts of quartz as much as 8 mm across. The iron-rich Gongo Soco quadrangle southwestward across the area laminae are 0.2-6.0 mm thick and consist of granules to the southwestern part of the Gongo Soco quadrangle and flakes of specular hematite intergrown with quartz, and northwestern part of the Conceigao do Rio Acima green chlorite, and chloritoid. The green chlorite and quadrangle (pis. 2, 3). The Sabara Formation is over­ chloritoid are minor constituents and occur mostly in turned and is overlain on the southeast by the Cerca­ the iron-rich laminae. The green chlorite occurs in dinho Formation, on the northeast it overlies the sheaves and plates curving about and between quartz Cercadinho Formation. Its southeast contact with grains and is crudely alined with the bedding planes the Cercadinho Formation is offset by a segment of the of the quartzite. Chloritoid is almost completely altered Fundao thrust fault in the northwestern part of the to limonite. It occurs as flakes and granules and replaces Conceigao do Eio Acima quadrangle and in the southern the quartz. Accessory minerals are zircon, titanite, part of the Gongo Soco quadrangle. and a micaceous mineral tentatively identified as stilpnomelane. UTHOLOGY

THICKNESS AND CONTACT RELATIONS In the Gongo Soco and Conceigao do Eio Acima quad­ rangles the schist and phyllite of the Sabara Formation The thickness of the Cercadinho Formation ranges have been deeply weathered to form saprolites mantled from about 400 meters in the western part of the Con- by a thin layer of lateritic soil. The schist and phyllite ceigao do Kio Acima quadrangle to 150 meters near form rounded topography of moderate relief, and the Lagoa das Antas and Lagoa do Couto. The thickness feldspathic quartzite, which is relatively more resistant along the south slopes of the Serra Geral ranges from to erosion, forms steep slopes and cliffs. Most exposures about 350 meters near the southwestern part of the of the weathered schist and phyllite are saprolite and Gongo Soco quadrangle to about 200 meters near N. were unsuitable for sectioning and microscopic study. 5,500, E. 10,000, in the eastern part of the Gongo Soco Also, because of deep weathering, faulting, folding, and quadrangle (pi. 2). Near the east boundary of the Gongo erosion of the upper beds, it was not possible to deter­ Soco quadrangle, the formation forms the axial zone of mine the stratigraphic distribution of the schist, phyl­ the Gandarela syncline and is displaced by faults; its lite, and quartzite in the Sabara Formation. thickness is not known. The phyllite weathers to reddish-brown, to light-gray The Cercadinho Formation, like the underlying for­ saprolite in which fine flakes of sericite and chlorite are mations of the Minas Series, is thicker in the southwest­ the only recognizable minerals. Some of the exposures ern part of the Gandarela syncline than in the north­ of saprolite have well-developed slaty cleavage. The eastern part. This parallel variation in thickness of the quartz schists weather to dark-red arenaceous clay in formations of the Minas has generally been considered which platy casts are the only remains of the micaceous to be the result of plastic flow of the rock along the minerals. more tightly folded parts of the Gandarela syncline Excellent exposures of feldspathic quartzite form low rather than of stratigraphic variations in the thickness cliffs that extend for more than 4 kilometers along a of the formations of the Minas Series. faultline scarp on the northwest slopes of the Serra da The contact between the Cercadinho Formation and Pedra Formosa from near N. 12,800, E. 2,880 (pi. 3), the overlying phyllite of the Sabara Formation is to N. 3,000, E. 8,000 (pi. 2). Weathered quartzites show poorly exposed in most localities; however, it appears to poorly developed foliation and are coarse grained, me­ be gradational through an interval about 3-5 meters dium gray, and schistose. Fresh quartzite is light gray, thick. Although the contact is covered at N. 3,300, E. and speckled with green epidote and mica. Thin sections 7,850 (pi. 2), the ferruginous quartzite of the Cerca­ show the rock to be feldspathic quartzite consisting of dinho Formation crops out within 3 meters of the over­ about 75 percent quartz, 15 percent feldspar, 5 percent lying phyllite of the Sabara Formation. Elsewhere in epidote, and 5 percent chlorite and sericite. Quartz and 124 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL feldspar grains are subrounded to angular, have crushed low undercut cliffs and ledges. Erosion has removed the boundaries, and range from about 0.02 to 1.0 mm across, softer underlying weathered rocks, leaving overhangs the median being about 0.5 mm. Most quartz and feld­ that break off, littering the slopes beneath the canga spar grains show considerable recrystallization and are cliffs. crudely alined parallel to the bedding of the quartzite. Originally the canga field in the mapped area prob­ A few larger metacrysts of feldspar lie at various ably covered a broad area of low relief between what angles to the foliation. Quartz is replaced along grain are now the ridges of Caue Itabirite that form the crests boundaries by epidote, sericite, and chlorite. The plagio- of the Serra da Pedra Formosa and the Serra Geral. clase is andesine, partly replaced by albite. Epidote oc­ Most of the canga was formed before the Tertiary up­ curs as aggregates of subhedral grains in the inter­ lift, when rejuvenation of streams rapidly eroded large stices of the quartz and feldspar grains. Chlorite and areas of the canga sheet downslope from the higher ele­ sericite occur as interbladed crystal aggregates and vations along the crests of the Serras Geral and da form folia about the quartz and feldspar grains. Acces­ Pedra Formosa. sory minerals are apatite and hematite; both are ran­ Most of the canga is common canga; it occurs along domly distributed throughout the rock. the ridges of Caue Itabirite and extends downslope from the crests as much as 1,000 meters. The layers of com­ THICKNESS AND CONTACT RELATIONS mon canga are enriched by leaching of carbonate and The original thickness of the Sabara Formation silica from the itabirite fragments and concurrent pre­ cannot be detemined because the upper beds have been cipitation of limonite cement. Enriched common canga removed by erosion and tight folding and plastic flow is widely used as a charge ore for the local charcoal have deformed the rocks in the axial zone of the Ganda- smelters because of its light weight and high porosity. rela syncline. A thickness of about 1,200-1,500 meters is Canga derived from weathering of Caue Itabirite in indicated for a partial section of the formation near the place occurs in irregular patches along the crests of southeast corner of the Gongo Soco quadrangle (pi. 2). ridges underlain by itabirite. The itabirite weathers along cleavage planes into partly collapsed and dis­ ST7RFICIAL DEPOSITS aggregated rock. The rock fragments are partly CANGA slumped but retain vestiges of the original attitude of Canga is the local name for a pavementlike surficial the itabirite bedrock. During the weathering process layer of iron-rich material that ranges from breccia- some of the silica and carbonate is leached out, and limo­ conglomerate, consisting of plates of itabirite and some nite is deposited along the fractures of the itabirite fragments of hard hematite cemented by limonite, to a fragments and minor amounts are disseminated and de­ fine-grained material made up of limonite cementing posited in the intervening itabirite fragments. The iron subordinate amounts of aluminous and ferruginous soils. enrichment may extend down into the unbroken Three types of canga are present in the mapped area: itabirite. Common canga, canga derived from weathering of Caue Chemical cangas are widespread in the Lagoa Grande Itabirite essentially in place, and chemical canga. Com­ Plateau area in the northwestern part of the ConceiQao mon canga consists of fragments of itabirite and minor do Rio Acima quadrangle. They are irregularly exposed amounts of hard hematite and other rock fragments in small patches by recent erosion that has stripped off cemented by limonite. The ratio of itabirite fragments soil, and presumably they comprise large areas of the to the limonite cement varies over a wide range, and Lagoa Grande Plateau a large field of canga deposited the iron content ranges from about 50 to 60 percent. on a deeply weathered erosional surface of dolomite. Canga derived from the weathering of Caue Itabirite This surface forms a gentle depression of low relief, essentially in place consists of disaggregated and partly interrupted by a few sinkholes, that is flanked on the slumped fragments of itabirite cemented by limonite. southeast by a ridge of Caue Itabirite and on the north­ Chemical canga (Simmons, 1960; Pomerene, 1964, p. 30) consists mostly of limonite cementing aluminous and west by a ridge of ferruginous quartzite of the Cerca- ferruginous soils. All types of canga may grade into all dinho Formation. The chemical canga grades southeast­ other types, saprolite, lateritic soil, or itabirite. ward into common canga that occurs above the itabirite Canga forms a hard resistant layer from 1 to 10 on the northwest slopes of the Serra da Pedra Formosa. meters thick and covers about 22 square kilometers of an Locally within the plateau area the chemical canga old erosion surface that is thought to be Tertiary in age. grades laterally into ferruginous lateritic soils and high- The canga is porous and very resistant to erosion. The alumina lateritic soil derived from the weathering of edges of the canga layer are in most places terminated by dolomite of the Gandarela Formation. ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 125 The chemical canga forms beneath a shallow soil cover The manganiferous laterite is a dark-brown to black and consists of very fine grained hydrous iron oxide plastic residual soil. It can be seen in quarry faces and cementing varying but minor amounts of pebble- to along stream bottoms as a 1- to 2-meter-thick layer that sand-sized rock fragments and ferruginous and alumi­ grades downward through a few centimeters into either nous lateritic soil. Chemical canga that has recently been nonplastic black saprolite or red manganiferous dolo­ uncovered by erosion characteristically has a botryoidal mite. The manganese in these laterites apparently surface, whereas weathered chemical canga has a rough resulted from the weathering of manganiferous dolo­ surface. mite, with attendant leaching of CaO and MgO from LATERITIC SOILS the dolomite and concentration of residual manganese, Ferruginous, aluminous, and manganiferous residual iron, and some silica. These laterites occur in areas of soils occur essentially in place as the result of sub­ low relief and limited circulation of ground water. tropical weathering of politic, dolomitic, .and igneous ALLUVIUM rocks. In areas of moderate or low relief these residual Recent deposits of unconsolidated gravel, sand, and soils form a 1- to 10-meter-thick layer of laterite over­ clay occur in the major stream valleys. Eemnants of lying and grading downward into saprolite that is as older alluvial deposits are found along the lower slopes much as 50 meters thick. of die valleys a& much as 30 meters above the present Ferruginous lateritic soils are widespread in these stream level. Most of the recent deposits consist of tail­ quadrangles and are derived from the weathering of ings from the extensive placer mining of the auriferous schist and phyllite. These laterites occur mostly at lower alluvial deposits during the 19th century. elevations and along the intermediate and upper slopes The alluvium consists of cobbles and pebbles of vein of the areas underlain by pelitic rocks. The laterite is quartz, quartzite, hard hematite, minor amounts of medium- to dark-red residual soil ranging from about itabirite, and schist in a matrix of clay and sand. 1 to 10 meters in thickness and averaging about 1-2 TALUS meters in thickness. Aluminous lateritic soils overlie granodiorite gneiss Talus accumulations are common at the base of cliffs. and, locally, dolomite of the Gandarela Formation. The Steep slopes are in many places covered with irregular granodiorite is easily eroded and forms low topography areas of material transported by soil creep, slump, and mantled by white, light-yellow, light-gray, and light- landslides. A major landslide near the Fazenda Gongo reddish-gray aluminous laterite consisting mostly of ,S6co consists of masses of schist and iron-formation in kaolin. Locally, where dolomite is exposed at higher crumbled schist and phyllite, saprolitic material, and elevations as on the Lagoa Grande Plateau where lateritic soil. drainage and circulation of .meteoric waters are good, IGNEOUS ROCKS the dolomite weathers to aluminous laterite consisting Igneous rocks of these quadrangles include granitic of a medium-reddish-brown low-grade bauxite soil. The and granodioritic gneiss, gabbro, and diabase. These bauxite occurs mostly as small irregular masses in the rocks are deeply weathered to saprolite mantled by canga field that caps the Lagoa Grande Plateau, and lateritic soil, and exposures of fresh rock are sparse. they grade laterally into chemical canga. These alumi­ Large areas of the Antonio dos Santos and Gongo nous laterites consist of either medium-reddish^brown Soco quadrangles are underlain by gneissic rocks of structureless dense earthy material or honeycombed granitic to granodioritic composition. The gneissic dense earthy material. A channel sample of typical rocks form only a small part of a regional batholith bauxite lateritic soil cut across a 6-foot width at that extends east, north, and west of the mapped area N. 9,500, E. 2,250, in the Conceigao do Eio Acima quad­ for many kilometers. The part of this regional bath­ rangle (pi. 3) contained 35.2 percent A12O3, 26.4 per­ olith that lies within this area is referred to as the cent 'SiO2, and traces of phosphorus, sulfur, and "Caete Complex", and this name will be used for the manganese. gneissic rocks underlying the Nntdnio dos Santos and Manganiferous laterite occurs along the projected strike of the upper beds of the Gandarela Formation. Gongo Soco quadrangles. Excellent exposures of these manganiferous! laterites Diabasic dikes have been intruded into thrust faults can be seen in the face of the Socorro quarry at N. 3,650, in the Cambotas Quartzite of the Serras das Cambotas E. 4,650 (pi. 2). They occur in a 300- to 400-meter-wide and Caraga. Plutons of gabbro and diabase intrude band extending from N. 4,200 on the west boundary the Cambotas Quartzite of the eastern part of the Serra of the Gongo Soco quadrangle to near N. 4,000, E. 6,000 das Cambotas and the gneiss of the Caete Complex (pis. (pi. 2). 1,2). 126 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL Dikes and veins of pegmatite are found in the schists CONTACT AND AGE RELATIONS of the Nova Lima Group at or near the contact with The contact between the gneiss and the schist, phyllite, the gneiss of the Caete Complex. Quartz-tourmaline and quartzite of the Rio das Velhas Series is obscured veins are also found at or near the contact of the gneiss in most places by deeply weathered saprolite, lateritic and the schist of Nova Lima age in the central part of soils, talus, and a thick growth of scrubby jungle; the Gongo Soco quadrangle. Quartz-kyanite veins however, the contact is well exposed in roadcuts and occur locally in the quartzite of the Maquine Group and railroad cuts in the central part of the Gongo Soco quad­ in the phyllite and ferruginous quartzite of the Cerca- rangle (pi. 2). dinho Formation. Milky quartz veins, fibrous quartz An excellent exposure of the contact between the veins, and quartz-hematite veins occur in the rocks gneiss and quartz-chlorite schist of the Nova Lima and metamorphic rocks of the area. Group can be seen in a roadcut at N. 6,800, E. 950 (pi. GRANITIC AND GRANODIORITIC GNEISS 2). Here the quartz-chlorite schist is in sharp contact DISTRIBUTION with gneiss saprolite. The foliation of the schist strikes west and dips 40°-50° S., and the contact strikes west Granitic and granodioritic gneiss of widely varying and dips 70°-80° S. A narrow layer of tourmalinized composition makes up most of the igneous rocks of the schist about 15 cm thick is in sharp contact with the Caete Complex north of the Serra Geral and along gneiss. Outward from this tourmalinized zone, the the west, north, and east flanks of the Serra das Cambo- schist is pervaded by clusters of tourmaline crystals tas in the Gongo Soco and Antonio dos Santos and veinlets of smoky quartz in a zone about 50 meters quadrangles. wide. This zone of disseminated tourmaline crystals The gneiss of the Caete Complex is easily weathered and smoky quartz veinlets grades over a few meters into and forms low rolling hills. Fresh exposures of gneiss a zone of pegmatite veins that is about 150 meters wide. are shown on plates 1 and 2 by patterned areas within the deeply weathered saprolite and lateritic soil. The contact between the gneiss and sericite-quartz schist and phyllite of the Nova Lima Group is well LITHOLOGY exposed at many places on the Caete-Barao de Cocais Although most of the gneiss is deeply weathered to road from near N. 8,500, E. 600, to N. 8,900, E. 10,800, laterite and saprolite, textures are well preserved in the in the Gongo Soco quadrangle (pi. 2). In these localities saprolite. The saprolite shows textures ranging from the contact is gradational from a gneiss saprolite equigranular to porphyritic. Fresh exposures of gneiss through a saprolite into weathered quartz- are white to light-gray medium-grained rocks consist­ sericite schist and phyllite. The width of the migmatized ing of plagioclase, potassium feldspars, quartz, bio- zone is variable and ranges from about 30 to 200 meters tite, sericite, and chlorite with granitoid textures that across. are faintly to moderately foliated. The contact between the Cambotas Quartzite and the Petrologic studies of thin section of fresh gneiss show gneiss along the west, north, and east flanks of the it to range in composition from granite to granodio- Serra das Cambotas is covered by lateritic soils and rite. The gneiss is composed of about 20-70 percent oligo- talus at the base of cliffs and steep slopes, and the con­ clase, 20-55 percent microline, 20-30 percent quartz, and tact was not directly observed. The gneiss cuts across 2-20 percent biotite by volume. the regional strike of Cambotas Quartzite near the Plagioclase grains are in part zoned and range from Serra do Garimpo, in the northeastern part of the Anto­ An10 to An25. A few of the zoned grains are selectively nio dos Santos quadrangle (pi. 1). Similar crosscutting altered along zones to sericite and rarely to epidote. relationships can be seen between the gneiss and the Some of the oligoclase grains are corroded by myrme- Cambotas Quartzite exposed within the main body of kitic quartz. Microcline and orthoclase make up varying the gneiss. amounts of the potassic feldspar. Microperthite inter- Near the contact between the Cambotas Quartzite and growths with the orthoclase are common. Quartz grains the gneiss, the quartzite contains irregular zones of dis­ are intergrown with the feldspars and form sutured seminated feldspar grains. These feldspathic quartzites boundaries with them in some sections; in other sections are believed to be alteration zones associated with an in­ the quartz and feldspar grains are broken. Most quartz trusive contact, rather than a sedimentary feature, be­ grains are severely strained. Biotite is greenish brown cause they do not occur within a definite stratigraphic and occurs as plates between grain boundaries of the interval. feldspar and quartz. Biotite is widely altered to aggre­ The intrusive contact between the gneiss and the gates of chlorite, epidote, and sericite, some of which quartz-chlorite schist of the Nova Lima Group in the are pseudomorphous after biotite books. western part of the Gongo S6co quadrangle is sharp. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 127 The gradational contact between the gneiss and the seri- A swarm of mafic dikes occurs in the Cambotas cite-quartz schists of the Nova Lima Group in the cen­ Quartzite of the Serra do Caraga and is intruded into tral and east-central parts of the Gongo Soco quad­ fault and joint planes. Along the east flanks of the Serra rangle is believed to be a zone of mobilized sericite schist das Cambotas a pluton of mafic rocks intrudes the Cam­ between gneiss and sericite-quartz schist. These sericite botas Quartzite. This pluton is about 4 kilometers long, quartz schists had a higher potassium content than the and only the western part of it lies within the mapped quartz-chlorite schist and may have been more easily area (pis. 1, 2). A few dikes and plutons of mafic rocks migmatized, and, therefore, the gneiss of the Caete occur in the granodiorite gneiss of the Caete Complex, Complex in the Gongo Soco and Antonio dos Santos and small segments of dikes and sills of mafic rocks are quadrangles is probably intrusive. poorly exposed locally within the pelitic, calcareous, and The age of the granitic and granodioritic gneiss of ferruginous metasedimentary rocks; however, they were the Caete Complex in the Antonio dos Santos and Gon­ not mapped, as they are deeply weathered and could not go Soco quadrangles is uncertain because the gneiss is be traced for more than a few tens of meters. found only in contact with metasedimentary rocks of The mafic dikes that comprise the swarm in the Serra the Rio das Velhas Series. Direct stratigraphic relations do Caraga are in most places weathered to reddish- with the rocks of the Rio das Velhas series indicate a brown saprolite or lateritic soil, but a few fresh pieces post-Rio das Velhas age. However, the schist, phyllite, of float were found in some of the outwashed alluvial and quartzite of the Rio das Velhas Series underlie rocks material. The dikes are easily mapped within the resist­ of Minas age that lie in the core of the Gandarela syn- ant Cambotas Quartzite, where they form shallow de­ cline and apparently were deformed by a post-Minas pressions along the crests of the ridges and deep ravines orogeny prior to the emplacement of the gneiss of the along the valley slopes. Caete Complex. This structural relationship indicates The weathered saprolite of the mafic dikes consists of that the age of the granitic and granodioritic gneiss of a reddish-brown clay matrix that contains pseudomor- the Caete Complex is post-Minas. phic phenocrysts of limonite after hornblende and The granitic rocks of the Quadrilatero Ferrif ero were pyroxene. The original fine- to coarse-grained porphy- thought to be Archean by early workers in the area. ritic textures of the dikes are partially preserved in the Barbosa (1954, p. 20), however, recognized that some saprolites. The borders of the dikes frequently have of the gneissic granitic rocks were derived from meta­ chilled zones ranging from a few centimeters to 15 cm sedimentary rocks of the Rio das Velhas Series. Herz in width and contain numerous pseudomorphs of hema­ and Dutra (1958) presented evidence for at least four tite and limonite after pyrite. ages of intrusive rocks in the Quadrilatero Ferrifero. The Cambotas Quartzite wallrock is apparently un­ The ages of the granitic were determined by altered by the dikes, except that it contains minor Prof. P. M. Hurley, Geology and Geophysics Depart­ amounts of disseminated cubes of pseudomorphous ment, Massachusetts Institute of Technology, who meas­ hematite after pyrite over a width of 5-10 cm. Numerous ured Ar40/K40 ratios in biotite. Four different ages were fibrous quartz veins occur in fractures in the dike rock determined: 2,500 m.y., 1,350 m.y., between 1,350 m.y. and the Cambotas Quartzite wallrock. and 500 m.y., and 500 m.y. The nearest dated granitic The mafic rocks of the large pluton exposed along the rock is in the east-central part of the Cocais quadrangle eastern part of the Serra das Cambotas also weathers in (Simmons, 1968) about 9 kilometers east of the Antonio most places to reddish-brown saprolite and lateritic soil. dos Santos quadrangle. Herz and Dutra (1957, p. 93) A few fresh exposures show the rock to be dark-green determined the age of a gneiss from the Cocais area to be medium-grained diabase consisting of plagioclase, pyroxene, and magnetite. Thin sections showed the rock 460 m.y. and correlated the gneiss of the Caete Complex to be made up of a calcic andesine and pyroxene matrix east of the city of Caete and west of the Serra das Cam- containing sparsely disseminated magnetite and ilmen- botas with the 500-million-year-old post-Minas granitic ite, and a few phenocrysts of pyroxene. The augite rocks. Guimaraes (1951, p. 37) also believed the granitic phenocrysts and the augite in the matrix are altered to gneiss of the Caete Complex in the mapped area to be an unidentified pleochroic amphibole. The magnetite is post-Minas in age. altered to limonite and the ilmenite to leucoxene.

MAFIC INTRUSIVE BOCKS QUARTZ VEINS Mafic igneous rocks occur in metasedimentary rocks of Quartz veins occur in all the metasedimentary and all ages and in the granodiorite gneiss of the Caete igneous rocks of the mapped area, but they are most Complex. conspicuously exposed in the more resistant rocks such 128 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL as quartzite, iron-formation, and feldspaithic quartzite. A pegmatized zone within the schist of Nova Lima The veins are of four types: milky (bull) quartz, the age can be seen in a roadcut near N. 6,800, E. 950, in the most common type; fibrous quartz; quartz-kyanite; and western part of the Gongo Soco quadrangle (pi. 2). The quartz-hematite. Quartz-tourmaline veins and smoky pegmatite veins occur in a zone about 200 meters wide. quartz veins are associated with pegmatite dikes and The contact between the granodiorite gneiss and the veins near the contact of granodiorite gneiss. schist is marked by a selvage of dark-green tourmaline Most of the quartz veins are composed of milky, or crystals. The schist beyond the selvage of tourmaline bull, quartz; a few clear quartz crystals occur as iso­ is cut by numerous veinlets and veins of smoky quartz lated aggregates in the milky quartz or as druses in and contains aggregates of disseminated tourmaline and vuggy parts of the veins. The veins occur in all the rock smoky quartz over a width of about 50 meters. This zone units but are most conspicuous in the Cambotas Quartz­ grades over a few meters into another zone about 150 ite and quartzite of Maquine age. The veins range in meters wide that consists of numerous quartz-feldspar- width from a few millimeters to about 50 cm. muscovite veins intruded into the foliation of the schist Quartz veins consisting of fibrous crystals alined at and into crosscutting joint planes. The veins range in right angles to the vein walls occur in association with width from about 4 cm to 4 meters. Some are discon­ the mafic dike swarms in the Cambotas Quartzite of the tinuous and form lenses parallel to the foliation of the Serra do Caraga. This type of quartz vein is particularly schist. abundant at N. 6,000, E. 10,000 (pi. 3). The veins range The pegmatite veins consist of quartz, potassic feld­ from about 0.5 to 5 cm across and crosscut the mafic spar, muscovite, and some actinolite. They are crosscut dikes and the Cambotas Quartzite wallrock. They are by smoky quartz veinlets, and some have borders of composed of thin, nearly parallel, matted rodlike crys­ smoky quartz. Most of the narrower (5-10 cm) veins are tals alined at right angles to the vein walls. The veins zoned and consist of coarsely crystalline clear quartz are in joint planes Ithat generally have a northerly trend and potassic feldspar borders and a central zone of mus­ and dip steeply to the east. covite and quartz. Some of the wider veins pinch out; Quartz-kyanite veins were observed in the Cercadinho and where they do, the vein consists of a central zone of Formation and in the Maquine Group but were not ob­ dark-gray acicular actinolite crystals and borders of served in the other rocks of the mapped area. highly fractured dark-brown to black quartz. Quartz-kyanite veins are exposed in a railroad cut The intervening quartz-chlorite schists within the in the Cercadinho Formation near the east boundary of pegmatized zone are deeply weathered to saprolite that the Gongo Soco quadrangle (N. 7,500, E. 13,300, pi. 2). preserves a well-developed migmatic texture. Numerous Here the edges of the veins are composed of milky lenses of smoky quartz occur along the foliation of the quartz with a central zone of felted blades of blue-green schist, and the minerals of the schist are moderately kyanite. Elsewhere debris from veins in the Cercadinho well crystallized into alternating dark- and light-col­ Formation can be seen, but exposures of veins are not ored bands. Migmatization of the schist is associated easily found. with the pegmatized zones and does not occur in the Quartz-kyanite veins are found at many places within schists above or below the pegmatized zone. the quartzite of the Maquine Group along the north­ METAMORPHISM west flanks of the Serra do Caraga (pi. 3) and consist Most of the Precambrian rocks of the mapped area of milky quartz enclosing large individual (2-5 cm have been subjected to a period of post-Minas low-grade long) laths of blue and blue-green kyanite. The kyanite metamorphism. Typical mineral assemblages, charac­ laths have ragged edges and are honeycombed by irregu­ teristic of the greenschist facies, are quartz, sericite, lar holes. chlorite, and chloritoid; and tremolite, carbonate, Quartz-hematite veins are common in the Batata! For­ phlogopite, and biotite. The Eio das Velhas Series may mation and in the Caue Itabirite of the Minas Series; have been subjected to an older period of metamor­ they occur as crosscutting veins about 1-5 cm across and phism; however, little evidence was observed in the consist of aggregates of quartz crystals intergrown with schists, phyllite, or quartzite that would suggest retro­ coarse curved plates of specularite. gression from a higher level of metamorphism. Pegmatite veins of quartz, feldspar, and muscovite Precambrian and impure sandstones of and of quartz and tourmaline are common in the migma- both the Eio das Velhas and the Minas Series have been tized and metamorphic quartz-chlorite schist and quartz- crystallized to quartzite and sericite, chlorite, and chlo- muscovite schist of Nova Lima age near the contact of ritoid-quartz schist. The iron-formations of both series the granodiorite gneiss of the Caete Complex. have been crystallized to siliceous and dolomitic itabirite ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 129 and minor amounts of tremolite and biotite. Shales, quartzite has been cut out by the Caraga thrust fault arenaceous shales, and volcanic rocks have been meta­ (pis. 3 and 4, sections A-A'-A", B-B'-B", and C-C'}. morphosed to phyllite and schist. The dolomite of the The northwest limb of the syncline is in schist of the Minas Series has been crystallized to dolomite marble, Nova Lima Group and quartzite of the Maquine Group. phlogopite, and talc. Feldspathic sandstones of the The contact between the schist and the quartzite is ap­ Minas Series have been crystallized to epidote, chlorite, parently discordant, but because of our poor exposures and sericite. the nature of this contact is not known. It may be an un­ Granitic gneiss of the Caete Complex has formed a conformity or a thrust fault. contact-metamorphic zone in schist and quartzite of the The exial zone of the Capivara syncline is most Rio das Velhas Series along the east and west flanks clearly defined near the south boundary of the Conceigao of the Serra das Cambotas and along the north slopes do Rio Acima quadrangle near N. 2,000, E. 6,200 (pi. 3). of the Serra Geral. The contact-metamorphic zone in Here quartzite beds double back in a closed pattern. The the schist ranges from about 100 to 300 meters in width axial zone was mapped northeastward between the and consists of pegmatitic dikes of quartz, feldspar, and northwest limb, which dips 55°-75° SE., and the over­ muscovite and quartz-tourmaline veins crosscuting ridden southeast limb, which dips about 25°-35° SE., to migmatized schist. Contact metamorphism of the near N. 6,000, E. 9,000 (pi. 3). Here, the axial trace quartzite of the Serra das Cambotas is indicated by swings abruptly eastward along the north face of the irregular zones of feldspar and specularite in the Serra do Caraga and the syncline is overturned to the quartzite that generally parallel the contact of the gra­ south (pi. 4, section A-A'-A"). The axial zone of the nitic gneiss. syncline is well exposed in readouts and in hogbacks of The granitic gneiss of the Caete Complex has been quartzite that are deformed by numerous drag folds deuterically altered by epidote, sericite, and chlorite. In along the north face of the Serra do Caraga. Both limbs most places alteration of the mafic rocks is unknown of the syncline dip steeply to the north (pi. 4, section because of deep weathering. The plutons and dikes of A-A'-A"}. gabbro and diabase do not have contact-metamorphic Along the west flanks of the Serra do Caraga the axial aureoles in the adjacent metasedimentary rock or gra­ zone of the syncline trends north and dips about 30° nitic gneiss. E. (pi. 4, section C-C'). Northeastward, along the STRUCTURAL GEOLOGY northwest face of the Serra do Caraga, the axis curves The Precambrian metasedimentary rocks of these to a northeasterly trend; and near N. 6,000, E. 9,000 (pi. quadrangles are deformed into three major northeast- 3), the dip steepens from about 30° to 85° SE. (pi. 4, section B-B'-B"). Eastward from this locality the trending isoclinal folds overturned to the northwest and broken by northeast-trending and generally south­ trace of the axis curves abruptly eastward, and along the north face of the Serra do Caraga the syncline is east-dipping low- to moderate-angle thrust faults. The overturned to the south (pi. 4, A-A'-A"). major elements of structural framework of these quad­ rangles are the Capivara syncline, the Conceigao anti­ CONCEICAO ANTICLINE ' cline, the Gandarela syncline, the Caraga thrust fault, The Conceigao anticline consists of a complexly folded the Fundao thrust fault, the Cambotas thrust fault, and belt of schist and iron-formation of Nova Lima age. The the Cambotas homocline. High-angle north-trending anticline is 4r-12 kilometers wide and extends across the east-dipping thrust faults occur in the upper plate of mapped area from the southwest corner to the east the Caraga thrust fault in the Serra do Caraga, and nu­ boundary; it is here named for the Conceigao valley, merous drag folds modify the major folds of the area. which roughly parallels part of the axial zone of this

FOLDS anticline (pi. 3). The anticline is a major feature in the mapped area and in the adjacent quadrangles. The CAPIVARA SYNCLINE structural details are largely obscured by deep weather­ In the southwestern part of the mapped area along ing of the schist, phyllite and iron-formation of Nova the west, northwest, and north flanks of the Serra do Lima age that make up the core of the anticline. Caraga, quartzite of the Maquine Group occurs as a The overturned northwest limb includes rocks of the canoe-shaped syncline overturned to the northwest. The Minas Series and the Nova Lima Group that dip about syncline is named for the nearby Fazenda Capivara. The southeast limb of the Capivara syncline is 30°-85° SE. The contact between the Minas Series and bounded on the west, northwest, and north by the the Nova Lima Group is discordant and may be either Caraga thrust fault. Cambotas Quartzite has been thrust an unconformity or a fault. The quartzite of the Ma­ over the southeast limb of the syncline, and part of the quine Group and the Cambotas Quartzite are missing 309-760 68 3 130 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL from the column along the northwest limb of the anti­ the western part of the Gongo Soco quadrangle and cline (pi. 4, section A-A'-A" and B-B'-B"}. The about 20°-35° SE. along the eastern part of the syn­ southeast limb dips about 30° E. in the southern part cline. In the vicinty of the Cabega de Ferro iron mine, of the Conceic.ao do Rio Acima quadrangle and about the northwest limb is deformed by drag folds and 35°-85° SE. along the northwest foothills of the Serra broken by a high-angle thrust fault. The southeast limb do Caraga (pi. 4, sections B-B'-B" and C-C'). The is broken by the Fundao thrust fault from the west southeast limb is made up of quartzite of the Maquine boundary to the east boundary of the mapped area. Group and Nova Lima rocks separated by a discordant The core of the syncline is in schist and phyllite of the contact. This discordant contact is poorly exposed and Sabara Formation, and the axial zone can only be gen­ may be either a fault or an unconformity. erally inferred because of poor exposures. The syncline In the southwestern part of the mapped area the is modified along both limbs by numerous small, open axial zone of this anticline plunges northeastward along cross folds. The syncline plunges southwestward in the a curving, north- to northeast-trending axis. The axial eastern part of the mapped area and northeastward in zone can be traced northeastward to near N. 10,000, E. the western part of these quadrangles (pis. 2,3). 8,000, in the Conceic.ao do Rio Acima quadrangle (pi. 3). Eastward from that locality the axial zone of the FAULTS Conceigao anticline is poorly defined because a series CARAQA FAULT of subsidiary folds is superimposed on the crestal zone The Caraea fault is a folded thrust fault that extends of the anticline. across the southeastern part of the Concei

An excellent exposure of the Cambotas fault is in a The displacement of the fault is not known. Movement roadcut along the Caete-Barao de Cocais highway on the fault seems to increase toward the northeast be­ (BR-31), near the east boundary of the Gongo Soco cause the Gandarela Formation and most of the Cer­ quadrangle (N. 9,150, E. 12,500, pi. 3). Here the fault cadinho Formation are cut out of the lower plate at N. strikes N. 70° E. and dips 40° S.; the fault zone con­ 4,700, E. 11,000 (pi. 3). tains 15-20 cm of gouge consisting of quartzite gran­ The Pedra fault is covered by canga and soil, but it ules and a shredded mica matrix enclosing angular is easily mapped along a zone of contrasting lithology fragments of vein quartz. The fault zone separates the and lateritic soils. It strikes about N. 50° E., and its cataclastic Cambotas Quartzite below from the schist of calculated dip is about 60°-70° SE. the Nova Lima above. The cataclasite and the schist Along the south slopes of the Serra Geral in the layers dip and strike nearly parallel to the fault zone, Gongo Soco quadrangle, a high-angle thrust fault in the but they are sheared and dragged into the gouge zone upper plate of the Cambotas fault (N. 7,600, E. 8,300, and are there obliquely truncated. At the south end of pi. 2) branches from the Cambotas fault and is traced the Serra das Cambotas the schist of Nova Lima age eastward along offset contacts of the Minas formations and the overlying rocks of Minas age in the north limb on the north limb of the Gandarela syncline. Dolomite of the Gandarela syncline have been thrust northwest­ of the Gandarela Formation is thrust northwest over ward over the homoclinal belt of north-trending east- Caue Itabirite along much of the trace of the fault east dipping Cambotas Quartzite that forms the Serra das of the village of Henrique Fleuisse to N. 8,450, E. 12,800 Cambotas. The overridden Cambotas Quartzite has (pi. 2). Here the upper plate of the thrust fault is been dragged and warped sharply to the west along the tightly folded; the dolomite of the Gandarela Forma­ eastern part of the Serra Geral (pi. 4, section A-A'- tion is cut out of the upper plate, and ferruginous A"), and at least 700 meters of quartzite has been cut quartzite of the Cercadinho Formation is thrust north­ out of the section by the overriding upper plate of Nova westward over Caue Itabirite. Lima and Minas rocks. Near the east boundary of the The fault trends about N. 70°-80° E., and its calcu­ Gongo Soco quadrangle the overriding rocks in the lated dip is about 65°-75° S. The throw is not known, north limb of the Gandarela syncline are deformed by except that it apparently increases eastward along the isoclinal folds overturned to the north and broken by a fault. The fault cuts the Gandarela Formation out of thrust fault. The isoclinal folds extend beyond the east the upper plate near the east boundary of the Gongo boundary of the Gongo Soco quadrangle for more than Soco quadrangle. 4 kilometers into the Santa Barbara quadrangle (G. C. In the Serra do Caraga a series of north-trending Simmons, written commun. 1959). east-dipping high-angle thrust faults cuts across and The magnitude of displacement along the Cambotas displaces open folds and an overturned syncline of thrust fault is not known with certainty; however, Cambotas Quartzite in the folded upper plate of the zones of cataclasite in the lower plate and large-scale Caraga thrust fault. The quartzite blocks moved from dragging and truncation of at least TOO meters of Cam­ east to west and overrode one another, imbricating the botas Quartzite along the south end of the Serra das upper plate of the Caraga thrust fault. These high- Cambotas and the east end of the Serra Geral indicate a angle thrust faults form the western part of a series of large displacement. imbricated slices of quartzite in the eastern part of the HIGH-ANGLE THRUST FAULTS Serra do Caraga. High-angle thrust faults occur along both limbs of These faults all dip about 70°-80° E. and trend north the Gandarela syncline and in the Cambotas Quartzite to northwest. Displacements along the faults are not of the Serra do Caraga and Serra das Cambotas. The known. Most of the faults are intruded by mafic dikes Pedra fault occurs along the southeast limb of the and are cut off by the Caraga thrust fault along the Gandarela syncline and can be traced from the northern northwest and north escarpments of the Serra do part of the Lagoa Grande Plateau (N. 13,700, E. 5,000, Caraga. pi. 3) along the northwest slopes of the Serra da Pedra West-trending reverse faults that dip steeply to the Formosa to near N. 5,000, E. 11,500 (pi. 2), in the south­ south are exposed along the crest and slopes of the ern part of the Gongo Soco quadrangle, where it is off­ set by the Fundao thrust fault. Northwestward from its southern part of the Serra das Cambotas in the north­ offset on the Fundao fault to N. 5,000, E. 11,400 (pi. 2), eastern part of the Gongo Soco quadrangle (pi. 2). The Caue Itabirite in the upper plate is thrust over dolomite magnitude of movement along these faults is uncertain; of the Gandarela Formation and ferruginous quartzite they displace a sequence of thick quartzite beds of Cam­ of the Cercadinho Formation. botas age and are truncated on the east by a pluton of ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIgAO DO RIO ACIMA QUADRANGLES 133 mafic rocks and on the west by granodiorite of the Caete and cross sections show only the surface contacts and Complex. their projection to a few thousand -meters below the CAMBOTAS HOMOCLINE surface, further inference of the structural features A north-trending and east-dipping belt of Cambotas may be shown by projecting the contacts above the Quartzite forms a homocline throughout most of the surface and to a greater depth. Serra das Cambotas in the northeastern part of the The projected contacts shown in section 1 are based mapped area. It is, bounded on the north, east, and west on the assumptions that (1) the Caraga and Cambotas by granodiorite and gabbro of the Caete Complex, and thrust faults are unconformable contacts; (2) the on the south by the Cambotas fault. Neither the top nor quartzites of Maquine and Cambotas ages that form the the bottom of the quartzite is exposed, except at the Serra do Caraga on the south and the Cambotas homo­ southwestern part of the homocline, where Cambotas cline on the north either were removed by one or several Quartzite unconformably overlies schist of the Nova periods of erosion or were not deposited throughout the Lima Group. It is not known whether the Cambotas central, eastern, and western parts of the mapped area; homocline is a tilted fault block, a limb of a fold, or an and (3) the Minas and pre-Minas rocks were then isoclinal fold. Zones of cataclasite, generally parallel to folded into three subparallel northeast-trending folds the bedding plane, occur throughout the homocline, sug­ overturned to the northwest. This reconstruction of the gesting widespread shearing within this belt of quartz­ structural evolution of the Precambrian rocks, if valid, ite. The relationship of the Cambotas homocline to the does not adequately explain why the schist of the Nova north-trending and east-dipping belt of quartzite in the Lima Group, the oldest known rock in the mapped area, central part of the Antonio dos Santos quadrangle is overlies younger Cambotas Quartzite along the south not known because of the intervening gneiss of the Caete slopes of the eastern part of the Serra Geral. Simmons complex. The western body of Cambotas Quartzite and Maxwell (1961) described this belt of schist as the within granodiorite gneiss probably is a roof pendant upper part of their Tamandua Group and interpreted and either a remnant of a faulted block or a limb of a the Cambotas fault as a minor bedding-plane thrust fold. fault. Dorr, Gair, Pomerene, and Rynearson (1957, p. SUMMARY 17-21), on the other hand, described a belt of schists in the Nova Lima quadrangle, the type locality, and The Precambrian rocks of the eastern part of the included them in the Nova Lima Group. The upper part Quadrilatero Ferrifero were folded and faulted during of this belt of schists was then mapped westward from several periods of orogeny. The correlation of the major the Nova Lima quadrangle (Gair, 1962) across the folds and faults, and to a certain extent of rock units Caete quadrangle (N. L. Alves, unpub. data) into the of pre-Minas age, is made difficult by the lack of precise western part of the Gongo Soco quadrangle. Thus, the structural and stratigraphic detail and by poor expo­ upper part of the Nova Lima Group as defined by Dorr, sures in much of the region. This difficulty in correlation Gair, Pomerene, and Rynearson (1957) and the upper and mapping has led to different interpretations of part of the Tamandua Group as defined by Simmons structural and stratigraphic features in the eastern part and Maxwell (1961) are apparently correlative in age. of the Quadrilatero Ferrffero. Perhaps the most con­ According to this interpretation, Cambotas Quartzite troversial subject is the discordant contact at the base and the quartzite of Maquine age along both limbs of the of the Minas Series and the discordant contacts separat­ Gandarela syncline were either removed by erosion or ing major rock units of the Rio das Velhas Series. In not deposited. The thickness of the Cambotas Quartzite the Cocais and Santa Barbara quadrangles (Simmons, in the mapped area indicates a rapid eastward thicken­ 1968) and in the Catas Altas and Capenema quad­ ing in the Serra do Caraga on the south and along the rangles (Maxwell, unpub. data) east and south of the eastern part of the Serra Geral on the north. Thus, it mapped area, these discordant contacts have been does not seem reasonable to assume that these thick clas­ mapped as unconformities. This interpretation was also tic units are absent from the stratigraphic column along followed by O'Eourke (unpub. data) in the Gandarela the eastern part of the southeast limb of the Gandarela quadrangle and by Alves (unpub. data) in the Caete syncline and along the eastern part of the northwest quadrangle. Structural features on and .along these con­ limb of the Capivara syncline because of nondeposition tacts are commonly associated with thrust faults. There­ or erosion. fore an alternate interpretation of the structure of the An alternate interpretation of the structural develop­ mapped area, and the interpretation currently followed ment of the Precambrian rocks of the mapped area is by Maxwell, Simmons, Alves, and O'Rourke is given in shown in section 2. It is based on the assumption that the figure 3, sections 1 and 2. Inasmuch as the geologic maps discordant contact between the basal formations of the 134 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL

B B' B" METERS 5000 -

4000 -

3000

2000 -

1000

0 -

1000

gdn 2000 -

3000

4000 - SECTION 1

SECTION 2 EXPLANATION

< Contact, dotted where projected K abov'e surface m _____ I < Fault, dotted where projected f-j above surface < I- tr ____ _ G. cc > Fold axis, dashed where projected 5 3 .!E T3 above surface < o O 2 u tr Minas Series D. Nova Lima Group, undivided 0 4 MILES _,______i______5 KILOMETERS

DATUM IS MEAN SEA LEVEL FIGTIEB 3. Inferred structural development in the Gongo S6co and ConceifiSo do Rio Acima quadrangles.. Minas Gerais, Brazil. Section 1 Inferred relationships of folded unconformable contacts to folds after post-Minas deformation. Section 2 Inferred relationships of folded thrust- fault contacts to folds after post-Minas deformation. Section based on section B-B'-B", plate 4. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 135 Minas Series and the schist of Nova Lima age, and the ous zone containing little gouge; in most places it is a Caraga and Cambotas thrust faults are all segments of clean cut surface separating the Minas iron-formation a regional thrust fault that developed during the early and phyllite from the underlying schist of the Nova stages of the post-Minas orogeny and was folded during Lima Group. Along the southwestern part of the south­ the later stages of the same orogeny. east limb of the Gandarela syncline, where the quartzite This projection assumes that a fault developed along of the Moeda Formation has overridden schist of the an initial break in the schists of the Nova Lima Group, Nova Lima Group, the quartzite is extensively sheared, ascended abruptly through the more competent and the overridden schists are deformed and crumpled. clastic units between the Minas and Nova Lima rocks, A reverse, but similar, relationship can be seen along the and then followed along the less competent beds of the north limb of the Gandarela syncline near the Camara basal formations of the Minas Series. Evidence sup­ gold mine, where Minas iron-formation has overridden porting this interpretation can be seen along the west, quartz schist of Nova Lima age. Here the overriding northwest, and north escarpments of the Serra do Ca­ iron-formation is deformed and crumpled, and the over­ raga, where the Caraga fault ascends westward through ridden schists are sheared and torn by numerous tear the overriding Cambotas Quartzite and the overridden fractures. quartzite of the Maquine Group. A similar relationship Along the north, northwest, and west escarpments can be seen along the eastern part of the Serra Geral, of the Serra do Caraga, the overriding Cambotas where the Cambotas fault descends eastward through Quartzite apparently dragged the quartzite of the Ma­ the overridden Cambotas Quartzite. Throughout the quine Group into an asymmetrical syncline overturned central, eastern, and western parts of the mapped area to the northwest, and the thrust sheet was emplaced by the fault followed along the basal beds of the Minas the forward movement of the upper plate and shorten­ rocks. Although the schists of Nova Lima age and the ing of the lower plate by folding. Gair (1962, p. 50) sug­ overlying basal rock of the Minas Series dip and strike gested that a large syncline of quartzite of Maquine nearly parallel to each other and to the contact that sep­ age in the Nova Lima and Kio Acima quadrangles arates them, the basal beds of the Minas rocks and the formed prior to the deposition of the Minas Series. Al­ schist of the Nova Lima Group are both obliquely trun­ though it is possible that the Capivara syncline was cated by this discordant contact. This indicates that the folded in pre-Minas time, regional trends of the iron- contact between the basal formations of the Minas Series formation and schist of Nova Lima age in the Con- and the schist of the Nova Lima Group is a fault. Simi­ ceigao anticline generally parallel the axial zones of the lar relationships can be seen along the Caraga fault in Capivara syncline and the Gandarela syncline in Minas the Serra do Caraga, where the overriding beds of Cam­ rocks, suggesting that all three folds were formed dur­ botas Quartzite are also obliquely truncated by the ing a single period of post-Minas deformation. Caraga fault. In the Serra do Caraga the Cambotas Quartzite of the IRON DEPOSITS upper plate of the Caraga fault is broken by a series of Iron ore is the most important mineral resource of north-trending east-dipping high-angle thrust faults, the area and includes high-grade hard and soft hema­ and the blocks of quartzite are sliced one over the other tite ore, intermediate-grade surficial canga ore, and from east to west, which suggests that the initial dis­ itabirite. placement of the thrust was toward the west in the Hard hematite ore contains about 66-69 percent iron. southern part of the mapped area. The abrupt warping It is low in silica, phosphorus, and sulfur and is widely of the south end of the quartzite of the Cambotas homo- used as charge ore for the local open-hearth furnaces, cline also suggests an initial east-to-west movement but none has been mined for export. along the Cambotas thrust fault in the central part of Oanga is the principal ore of intermediate grade and the area. The forward movement of the thrust appar­ contains about 50-60 percent iron. Although canga has ently was followed by folding of all the Precambrian a fairly low and variable iron content and a fairly high rocks into three northeast-trending folds overturned to phosphorus content (0.12-0.18 percent), it is widely the northwest. used in the local charcoal blast furnaces because it is The fault zone in places where quartzite has overrid­ light in weight and porous and is therefore easily den quartzite consists locally of breccia zones, but in smelted. most places it consists of granular quartzite and mica. Iron content of itabirite varies widely, ranging from Along the Gandarela syncline where the basal phyllite about 25 to 50 percent and averaging about 40 percent. and iron-formation of the Minas Series have overridden The deeply weathered itabirite is soft and friable in schist of Nova Lima age, the fault zone is an inconspicu­ most places. It can be easily mined, is amenable to con- 136 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL centration, and is a source of very large tonnages of posed on the north limb of the Gandarela syncline. material that can be concentrated to a 65-67 percent Although this anticline is isoclinal at the east boundary iron product. Itabirite is enriched beneath the wide­ of the quadrangle, it becomes open about 1,000 meters spread layer of canga, in most places, to as much as 55 west of the mine area and then becomes parallel to the percent iron and is locally mined along with the canga regional dip and strike of rocks in the north limb of the ore. Gandarela syncline. This westward structural transi­ HIGH-GRADE HARD HEMATITE QBE tion from an isoclinal anticline through an open anti­ High-grade hard hematite ore occurs as irregular cline to a regional trend is accompanied by westward replacement bodies in the itabirite and dolomite of the diminishing of the replacement of the dolomite by hema­ Itabira Group. Hard hematite ore occurs in the Caue tite in the core of the anticline. This relationship sug­ Itabirite at the Piao prospect and at the Santo Anto­ gests that folding was the primary control for the nio mine near the west boundary of the Conceigao do Rio emplacement of the hard hematite ore body exposed in Acima quadrangle (pi. 3). It occurs in dolomite of the the Cabega de Ferro iron mine. Gandarela Formation in the Cabega de Ferro mine The hard and soft hematite ores at the Piao prospect near the east boundary of the Gongo Soco quadrangle are in the lower part of the Caue Itabirite. Both retain (pi. 2). the original laminations of the itabirite, which dip and The hard hematite ores are composed dominantly of strike nearly parallel to the enclosing hard siliceous very fine grained specularite with only very minor itabirite (fig. 4). The contact between the hard hema­ amounts of gangue minerals, which include quartz, talc, tite ore and the hard itabirite is variable, ranging from and chrysotile. Hard hematite ore in the Caue Ita­ sharp to a gradational zone as much as 4 meters wide. birite contains only a few widely spaced tiny blebs of Individual laminae of the hard itabirite grade through quartz that are visible on weathered surfaces. Other­ a narrow zone of enriched itabirite into hard high-grade wise the ore appears to be pure hematite. Gangue min­ hematite ore with no apparent volume change. The con­ erals are more abundant in the replacement bodies of tact between the softened platy hematite ore and the hard hematite in the dolomite of the Gandarela For­ hard itabirite is concealed, but in a few places the soft mation and occur as layers and lenses of talc and ore crops out with ledges of hard hematite, and it is chrysotile. apparent that the contact between the soft ore and the Harder and Chamberlin (1915, p. 395) suggested itabirite is either sharp or gradational over a few that the high-grade hematite ore bodies were deposited meters. as iron-rich layers in the itabirite. Dorr, Guild, and The crosscutting nature of the hard and soft hematite Barbosa (1952, p. 295) observed that hard hematite ores and the presence of the well-preserved laminae in ore, in many places, crosscuts the laminations of the the high-grade hard and soft ores indicate that the itabirite, and they presented evidence that the high- ores formed !by metasomatic replacement of itabirite by grade hematite ores were formed by metasomatic re­ iron-bearing fluids. The source of the iron apparently is placement of quartz and carbonate of the itabirite by the enclosing itabirites, which are poorly laminated, iron-bearing fluids. Guild (1957, p. 55) suggested that unusually low in iron (25-30 percent), and very hard. brecciated zones in the itabirite in the Congonhas dis­ The hard itabirite enclosing the ore forms hogbacks trict acted as channels for iron-bearing fluids that re­ extending southwest and northeast from the ore body placed the quartz and carbonate with hematite. Dorr for more than 1,000 meters. The itabirite was apparently and Barbosa (1963, p. 73) believed that the high-grade indurated by through-going fluids that mobilized the hematite ore bodies in the Itabira district were localized iron and concurrently deposited silica that was later along the axial zones of folds because of the rela­ replaced by hematite in the ore body. tively lower pressure and attendant relatively higher The hard and soft ores of the Piao prospect lie within porosity there; these conditions facilitated the flow an area of regular laminations in the itabirite and hard of iron-bearing fluids that replaced the quartz of the hematite, and there is no obvious structural control for itabirite. the localization of ore. Hard hematite ore bodies in the mapped area crosscut PlXo PROSPECT the lamination of the itabirite and dolomite of the Ita­ The Piao prospect is at the west boundary of the Con- bira Group and formed either during or after the de­ ceigao do Rio Acima quadrangle (N. 4,700, E. 100, pi. 3) formation and metamorphism of the Precambrian approximately 2 kilometers southwest of the Fazenda rocks of the region by post-Minas orogeny. da Mutuca in the Gandarela quadrangle. The prospect The hard hematite ore body exposed in the Cabega de is accessible by a road that leads southeastward from the Ferro mine is in the axial zone of an anticline superim­ village of Rio Acima through the village of Gandarela ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO EIO ACIMA QUADRANGLES 137 to the Fazenda da Mutuca. From the fazenda a trail of a crude U-shaped body of high-grade hard hema­ leads southeastward to the Piao prospect. tite bounded on the east, north, and west by high-grade The Piao ore body consists of a prominent tor of hard platy, partly softened hematite and soft powdery hema­ hematite on the northwest slopes of a small valley. It tite, and on the south by hard siliceous itabirite. The is flanked on the northwest by a ridge of Oaue Itabirite hard hematite grades over a few meters into the platy and on the southeast by a ridge of phyllite and quartzite hematite, which apparently is partially disaggregated of the Cacaga Group (fig. 4). The tor of hard hematite by weathering. The hard and soft hematite and the dis­ extends above the general slope of the valley to a height aggregated platy hematite grade laterally over a few of about 70 meters; it averages about 200 meters across meters through partially enriched itabirite into poorly at the base and about 30 meters across at the top. About laminated iron-poor (25-30 percent iron) hard itabirite. 2 million tons of high-grade hard hematite is exposed The hard hematite consists of very fine grained spec- above the level of the valley slope. The computation of ularite with only a few blebs and druses of quartz de­ reserves is based on the volume of a cone 70 meters high posited in small vugs that form along the relic laminae and 200 meters in diameter and an assumed 4 metric tons preserved in the ore. These vugs are sparsely distributed of ore per cubic meter. Projected ore reserves below the and are about 0.5-1.0 mm wide and about 3-6 mm long. present erosional surface are 80,000 tons of high-grade Laminations of the replaced itabirite are well preserved hard hematite per meter of depth. At the time of map­ as relic textures in hard hematite. The relic laminations ping no ore had been mined from the Piao ore body. are regular, trend N. 60° E., and dip steeply to the south­ The Piao ore body is stratigraphically within the east or northwest. The hard hematite is cut by anasto­ lower 200 meters of the Caue Itabirite. The ore consists mosing linear shatter zones in which fragments have

EXPLANATION iiiiii x^x^x^y^joAyx-x-x-xvr:^*; ^^^^^^m^^^y

Caue Itabirite hh, hard hematite ^^^^^^^^^^^^^^ sh, soft hematite

Moeda Formation

Approximate contact Geology by S. L. Moore, 1956 80 80

Inclined Overturned Strike and dip of beds

0 50 100 200 300 METERS CONTOUR INTERVAL 50 METERS DATUM IS MEAN SEA LEVEL

FIGURE 4. Geologic map an'l section of the Piao prospect, ConceicSo do Elo Acima quadrangle. 138 GEOLOGY AND MINERAL RESOURCES OP PARTS OF MINAS GERAIS, BRAZIL been only slightly rotated. The fragments range from No complete record of the ore mined was available to about 1 to 20 cm across and are poorly cemented with the writer, but the size of the three major open pits hematite, limonite, and minor amounts of quartz and indicates that several tens of thousands of tons of hard manganese oxide. The fractures in the shattered zone hematite ore has been mined. The hard and soft hema­ fade out laterally into massive hard hematite over a tite ore is exposed in an area approximately 200 meters zone of about 10-20 cm. A few joints cut the massive long and 200 meters wide; the projected reserves of hard hematite and trend N. 40°-45° W. and have a near ore containing more than 65 percent iron are about 150,- vertical dip. 000 metric tons of hard and soft hematite ore per meter The hard hematite of the Piao ore body is bounded of depth. The grade of ore is variable, however, .because on the east, north, and west by an irregular mass of of incomplete replacement of phyllite and talc layers high-grade soft powdery hematite and by partly soft­ in the dolomite and varying amounts of vein chrysotile ened and disaggregated hard hematite consisting of in faults and joint planes. a well-laminated mass of hematite plates. Both types The hematite ore exposed in the Cabega de Ferro of ore are apparently derived from the weathering of mine is a replacement of dolomite of the Gandarela the hard hematite ores. They grade laterally over a few Formation in the core of an eiast-plunging isoclinal an­ meters into one another and into the hard hematite of ticline overturned to the north. Bedding-plane features the ore body. On the east side of the central mass of of the dolomite are preserved in hard and soft hematite hard hematite, excellent exposures of partly weathered ores in the form of talc laminae, phyllite partings, and hard hematite crop out as a partly disaggregated mass fibrous chrysotile parallel to laminae in hematite, and of platy hematite. The plates of hematite are separated by vugs that parallel the relic laminae of the hematite by thin laminae of finely divided flakes of specular ore. The bedding in the core of the anticline trends hematite that formed as the result of weathering parallel N. 59° W. to N. 70° E. and dips 20°-45° SSE. The ore to the relic laminations of the metasomatically replaced body is flanked on the north by an east-plunging iso­ itabirite. clinal syncline of ferruginous quartzite, and on the The soft hematite ore is mostly covered by soil and southeast by ferruginous quartzite, both of the Cerca- talus; however, recent erosion has exposed irregular dinho Formation. The ore body is localized mostly on areas along the base of the hard hematite. The soft ore the axis and the north limb of the anticline. This iso­ is characteristically a bluish-gray noncoherent mass of clinal anticline and the isoclinal syncline both were finely divided plates of hematite containing some frag­ traced east from the mapped area for more than 4 kilom­ ments of hard hematite. eters in the Santa Barbara quadrangle by Simmons A chip sample taken from the base of the tor of hard (1968). Westward from the mine area the isoclinal anti­ hematite at 2-meter intervals assayed 69.6 percent iron, cline rapidly becomes asymmetrical in form, then flares 0.03 percent phosphorus, and traces of manganese, into an open fold, and then becomes part of the north sulfur, and SiO2. limb of the Gandarela syncline about 1,500 meters west- southwest of the mine area. This westward structural CABECA DE FERRO MINE transition from an isoclinal anticline to an open anti­ The Cabega de Ferro mine includes three open pits cline is accompanied by a decrease in the grade of ore, and several smaller pits and prospect pits in hard and which suggests that only the more tightly folded part soft hematite near the east boundary of the Gongo of the dolomite in the anticline was metasomatically Soco quadrangle (N. 8,000, E. 13,000, pi. 2); other open replaced. pits in hard hematite are located in the Santa Barbara The ore of the Cabega de Ferro mine consists of al­ quadrangle, which joins the Gongo Soco quadrangle on ternating layers of hard and soft hematite, a few inter- the east. Only the deposits in the Gongo Soco quad­ beds of phyllite, and talc partings. The hard hematite rangle are described. The mine workings are accessible retains the original bedding laminae of the dolomite, by a road leading west from the town of Barao de Co- cais in the Santa Barbara quadrangle. and numerous small lenticular vugs 1-5 mm in length Iron and manganif erous iron ore have been mined in­ parallel the relic laminae in the hard hematite. Soft termittently from the Cabega de Ferro mine for more hematite layers consist of irregular partings of finely than 20 years by the Companhia Brasileira de Usinas divided flakes of specularite and separate layers of hard Metalurgicas of Barao de Cocais. Although ferrugin­ hematite. A few partings of dark-gray ferruginous ous manganese ore has been extracted from some parts phyllite and laminae of light-gray talc and white chrys­ of the mine, the ore is principally hard and soft high- otile are interlayered with the hard and soft hematite. grade hematite. The ore is screened, and the lump ore The hematite ore along the crest of the fold is broken is used for charging open-hearth furnaces. by numerous northeast-trending southeast-dipping ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEigAO DO RIO ACIMA QUADRANGLES 139 joints and minor faults, and some of these joints and high-grade hard hematite and itabirite. Soft ore in most faults are filled with an opalescent fibrous chrysotile places contains various amounts of hard hematite and alined at right angles to the fractures. Chrysotile also grades through platy or schistose partly coherent hema­ occurs as partings in the ore. The fibers are alined paral­ tite ore into hard hematite. The intercalated layers of lel to the bedding plane and form herringbone patterns. soft ore generally parallel the bedding and foliation, but Analyses of eight channel samples cut vertically from in places they crosscut the bedding and foliation and oc­ the faces of pits 1,2, and 3 of the Cabega de Ferro mine cur as irregular bodies enclosed by itabirite. are shown in table 3. The soft hematite ore was derived from the hard SANTO ANTONIO MINE hematite ore by supergene leaching of hematite along The Santo Antonio mine is at N. 6,080, E. 1,250 (pi. grain boundaries, which produced either platy partly 3), in the Conceigao do Kio Acima quadrangle on the coherent hematite or soft micaceous hematite ore. east slopes of the Corrego da Gandarela valley. The Soft platy hematite layers are common in the Caue mine can be reached by trails leading southeastward Itabirite in an area about 3 kilometers long extending from the Fazenda da Mutuca along the Corrego da from the Gongo Soco iron mine west to the western Gandarela. The deposit was prospected by J. Morgan part of the Gongo Soco gold mine area. At many places da Costa of the village of Kio Acima, and about 500 tons along this belt of itabirite, surface exposures contain of high-grade hard hematite ore was transported to the numerous lenses and layers of soft hematite ore inter­ Fazenda da Mutuca by mules and thence to Kio Acima calated with some hard hematite layers, and poorly by truck. laminated itabirite. These replacement layers of soft ore The mine consists of two open pits cut into two 5- and some hard layers occur in itabirite that has been meter-thick layers of hard hematite. The hard hematite highly deformed by drag folds and broken by closely layers are interbedded with soft siliceous itabirite and spaced shear planes. The itabirite between the layers are bedding replacements in itabirite. The exposed of soft ore is poorly laminated because the carbonate length of the layers of hard hematite is about 50 meters. and quartz of the light-colored laminae have been partly There are no surface exposures beyond the ore laid open replaced by hematite and the laminae, thus, obscured. by the open pits, and the true horizontal extent of the Most of the intercalated lenses and layers of soft ore body apparently is about 50 meters. and hard hematite occur in dolomitic iron-formation in The hard hematite layers are interlayered with lami­ the middle and upper parts of the Caue Itabirite. Ac­ nae of soft, powdery, finely divided flakes of specular cording to Hen wood (1871), the itabirite and the soft hematite. The hard hematite ore is broken by numerous hematite layers contain pockets of metallic gold, and the fractures into rhombic fragments that are about 2 cm itabirite apparently contains small amounts of fine gold. across or more. The original itabirite laminae are well The richer pockets of gold occur in "jacutinga" 2 along preserved in the hard hematite layers. drag folds. The surface exposures of soft hematite lay­ Analyses of two channel samples cut across both lay­ ers occur in zones of highly deformed itabirite; ap­ ers of hematite show an average of 69.7 percent iron, parently soft and hard layers of hematite replaced tlia 0.4 percent SiO2, 0.20 percent manganese, 0.02 percent itabirite in areas that were highly deformed. phosphorus, and traces of A12O3. The source of the iron for the soft and hard iron- ore replacement layers near the Camara and Gongo HIGH-GRADE SOFT HEMATITE ORE Soco gold mines may have been through-going iron- High-grade soft hematite ore occurs around the west, bearing fluids that dissolved iron from the adjacent north, and east flanks of the Piao ore body, and as re­ itabirite, or the iron could have been precipitated from placement layers in the Caue Itabirite intercalated with 2 See discussion of the term "jacutinga" on page 145. TABLE 3.< Analyses (in percent) of hard hematite ore from the Cabe$a de Ferro mine [Analyst; Dr. Cassio Pinto, Departmento Nacional da Producao Mineral, Belo Horizonte]

i la Ib 2 2b 2c 2d 4 4b

Fe.._. ______._____.___._._._.__ _....._____. 68.5 68.4 68. 5 68.4 68.5 68.0 68.4 68. 1 67.3 SiO, ______. 9 1.0 . 8 1.3 1.0 1.4 1.4 1.8 1.6 AlaOs ------___--_-.____ .5 . 5 . 6 . 5 . 5 . 5 . 5 . 5 .6 Mn _ . 4 . 5 . 5 . 5 . 5 . 6 . 2 .3 .4 P_____.._ . 02 .02 .02 . 02 .02 . 02 . 02 . 02 .02

1. la, Ib. Channel samples taken from open pit 1. 2. 2b, 2c, 2d. Channel samples taken from open pit 2. , 4t>. Channel samples taken from open pit 3. 140 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL

hydrothermal fluids emanating from the granitic gneiss Excellent exposures of hard siliceous itabirite crop of the Caete Complex. out along the crest of the Serra da Pedra Formosa near The soft hematite ores of the Gongo Soco mine could the west boundary of the mapped area. These hard sili­ be cheaply mined, the hard hematite fraction used for ceous itabirites enclose, and extend outward from, the charge ore in open-hearth furnaces, and the soft ore high-grade hard and soft hematite of the Piao ore body. used as sintering ore for charging charcoal blast fur­ The hematite and the quartz of the hard itabirite are naces. It also may be possible to recover the disseminated poorly segregated, the laminae are indistinct, and the metallic gold in the ore. itabirite contains relatively little iron (25-30 percent). ITABIRITE The writer believes that these extensive exposures of Soft siliceous itabirite and hard dolomitic and sili­ hard siliceous itabirite near the Piao ore body are re­ ceous itabirite are exposed in the mapped area. However, sistant to weathering because they have been altered by most of the itabirite along the crest and slopes of the through-going fluids that mobilized part of their iron Serras Geral and da Pedra Formosa and beneath the content and concurrently precipitated silica which widespread canga field is soft noncoherent siliceous changed original crystalline textures. itabirite. Exposures of hard dolomitic itabirite crop out lo­ The soft itabirite could be easily mined by open-pit cally as small ledges at many levels in the upper and methods without drilling, blasting, or crushing because middle parts of the Caue Itabirite along both limbs of it is deeply weathered in most places to a noncoherent the Gandarela syncline. Hard dolomitic itabirite is pres­ mass of granular quartz and hematite that can be easily ent along the base of a canyon cut through the Serra da concentrated to a product of more than 65 percent iron Pedra Formosa by the Corrego da Gandarela. The hard (Dorr and Barbosa, 1963, p. 54). The soft itabirite con­ dolomitic itabirite can be traced upward along near- tains about 35-45 percent iron, and the hard dolomitic vertical laminations from the base of the canyon into and siliceous itabirites contain about 25-35 percent iron. softened siliceous itabirite that forms the upper walls of At present the soft itabirite is a source of vast tonnages the canyon and the crest of the serra. The relationship of concentrating material and constitutes a valuable re­ of the hard dolomitic itabirite to the present erosional serve of hematite for future use. surface and the numerous isolated exposures of hard Most geologists who have worked in the Quadrilatero dolomitic itabirite in the middle and upper parts of Ferrifero generally agree that the soft siliceous itabi­ the Caue Itabirite suggest that much of the itabirite of rites have been disaggregated by ground waters. There the mapped area contained carbonate. is some question, however, regarding the chemical na­ The writer believes that presence of carbonate in the ture of the ground water, how it disaggregates the ita­ itabirite and physiographic position are the major fac­ birite, and the mineralogic composition of the original tors in softening of the itabirite. Siliceous itabirite may unaltered itabirite. also disaggregate through the leaching of silica, as does Guild (1957, p. 45) and Pomerene (1964, p. 44) be­ the dolomitic itabirite, but carbonate-bearing itabirite lieved that soft siliceous itabirite once contained small apparently is more easily disaggregated by meteoric amounts of carbonate that was easily dissolved, along waters because the carbonate is easily dissolved, partly with some silica, by meteoric water; and as the meteoric disaggregating the itabirite, and the alkaline waters water became more alkaline, it further increased its continue to dissolve silica at an accelerated rate, com­ ability to dissolve silica along the boundaries of the pletely disaggregating the itabirite. quartz grains and thus disaggregated the itabirite. Dorr Many small exposures of hard dolomitic and siliceous and Barbosa (1963, p. 26-27), however, believed that itabirite are either enclosed by or intercalated with soft most of the soft itabirite of the Itabira region was once siliceous itabirite; thus, for these small exposures physi­ siliceous itabirite that was disaggregated because of its ographic position and the absence or presence of carbon­ physiographic position, textures, and duration of ex­ ate apparently have not been factors in weathering of posure to surface weathering. They believed that the the itabirite. In many places the hard itabirite is not hard siliceous itabirite in the deeply incised canyon in the Itabira area has retained essentially the original near ore bodies or in zones of structural compression. crystalline texture of itabirite after it was metamor­ It apparently remains hard because of differences in the phosed during the post-Minas orogeny. crystallization of the chert, carbonate, and iron minerals Although both hard dolomitic and siliceous itabirite during the post-Minas orogeny. Where hard siliceous crop out locally, most of the itabirite of the mapped area itabirite encloses or occurs near hard and soft hematite has been weathered to partly coherent granular quartz ore bodies, as at the Piao prospect, the original crystal and hematite. textures of the itabirite probably were changed by ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEICAO DO EIO ACIMA QUADRANGLES 141 through-going fluids that deposited the replacement through a 1-cm screen, and only the coarser fractions hematite. are sent to the charcoal blast furnaces. CANGA The composition, origin, and methods of mining of Ore-grade canga in these quadrangles has been mined the canga ores make it difficult to generalize the average only from common canga made up mostly of fragments grade and physical characteristics of the ore. However, of itabirite and a few fragments of hard hematite and the average composition of 142,908 tons of canga shipped schist. Chemical cangas are of ore grade, but most are from the Cabral mine to the steel plant of the covered by soil and are hard and heavy; they are there­ Companhia Siderurgica Belgo-Mineira from 194T to fore not suitable for use in the charcoal furnaces. Canga 1952 was 58.25 percent iron, 7.31 percent SiO2,1.71 per­ that forms in place is generally low grade, occurs only cent A12O3, 0.16 percent phosphorus, and 0.47 percent in small areas, and is not suitable for use in the local manganese. Canga contains varying amounts of iron furnaces. and has a high phosphorus content, but it is widely used Common canga often has a higher iron content than as charge ore in the local charcoal smelters because, the itabirite from which it is derived because of both owing to its light weight, it will not crush the charcoal. natural mechanical concentration and chemical enrich­ Canga covers an area of approximately 22 square kil­ ment. The weathered and softened itabirite fragments ometers within the mapped area. Total reserves, assum­ are partly disaggregated during transport by water or ing a 1-meter thickness and three tons of ore per cubic by downslope migration, and some of the quartz and meter, are about 66 million tons of material ranging carbonate granules are separated from the itabirite frag­ from about 50 to 60 percent iron. Large areas within the ments, upgrading the iron content of the fragments. canga sheet are unsuitable for use as ore, however, be­ In places where high-grade hard hematite fragments cause canga grades laterally into lateritic soils and are deposited along with itabirite fragments, the iron itabirite. On the other hand, many areas in the canga field are more than 1 meter thick. Also, in most areas content of the detritus is further increased. The itabirite beneath the canga there is a 1- to 3-meter layer of en­ and hard hematite fragments are then cemented by riched itabirite that can be mined along with the canga limonite precipitated from iron-bearing meteoric ore and which would increase the total tonnage of ore waters. between 55 and 60 percent iron that could ultimately be Although the thickness of the canga sheet ranges mined. from about 1 to 10 meters and average about 1.5 meters, Because of the irregular thickness and extent of the all the canga mines in the mapped area, which include canga layer and the unmarked boundary lines of the the Gongo Soco, Barra Mansa, Cabral, Trindade, and mines, reserves have not been estimated for the mines Ponte da Paixao, are developed on a 1- to 3-meter-thick discussed below. layer of common canga overlying Caue Itabirite. The iron content of the canga ranges from about 55 to 60 per­ CONGO SOCO MINE cent ; however, at all the mines, about 1-2 meters of the The Gongo Soco iron mine (pi. 2) is about 1 kilo­ underlying enriched Caue Itabirite is removed along meter north of the village of Venda Velha. The mine with the canga layer. The enriched itabirite is somewhat workings are bisected by the Central do Brasil railroad, lower in grade than the canga and contains about 50-55 which leads east from Belo Horizonte to . The percent iron. area is also accessible by a road that leads south from The depth of enrichment of the itabirite beneath the the Caete-Barao de Cocais road. Canga and enriched canga is quite variable in the mine workings; it ranges itabirite have been mined from this area since 1937 by from about 0.5 to 3 meters and averages perhaps 1.5 Companhia Ferro Brasileira for reduction in the Caete meters. Some of the quartz and most of the carbonate smelter. The grade of ore shipped and the total produc­ have been leached by downward circulation of meteoric tion of this mine from 1937 to October 1956 are shown water, and limonite has been precipitated so that the in tables 4 and 5. itabirite beneath the porous layer of canga is enriched. The mine has been developed along a random series The common canga and the underlying enriched itab­ of benches over an irregular area 500 meters long by irite have been mined by hand methods in most of 400 meters wide, on both sides of the Central do Brasil the mines. Blocks of the canga layer are spalled off with railroad. The (benches are cut to the base of the enriched hand tools from an undercut ledge and broken into zone in the underlying itabirite. The height and width smaller pieces with sledge hammers; the canga frag­ of the bench face are correspondingly greater where the ments and the itabirite fragments are then screened depth of enrichment is greater. 142 GEOLOGY AND MINERAL RESOURCES OP PARTS OF MESTAS GERAIS, BRAZIL TABLE 4. Analyses (in percent) of canga ore and enriched itabirite Caete to Barao de Cocais. The mine was opened in 1935 ore from the Gongo Sdco mine by the Companhia Siderurgica Belgo-Mineira. Produc­ [Analyses represent bulk samples taken from the ore storage bins of the Companhia Ferro Brasileira smelter at Caetfi, Minas Qerais, 1956. Analyst unknown] tion of canga and enriched itabirite ores began in 1947 and continued until 1953, when mining was temporarily Date Fe SiOa Mn AljOj H2O Loss on ignition suspended. The grade and tonnage of ore extracted from the Cabral mine are given in table 6. April 4.... __ . ____ 61.90 5.2 0.30 0.30 4.0 2.8 The mine was worked as a series of benches cut into June 13 .. _____ 55.17 15.6 .16 1.11 13.0 2.7 25.. __ . _____ 55.70 11.5 .14 1.30 9.0 4.7 the south slopes of the Serra Geral. The benches have 30 _____ . ____ 58.20 8.9 .17 1.10 6.5 4.5 exploited an area of canga about 600 meters long and July 6..- 57.10 11.0 .10 1.00 8.0 4.5 20. ______58.20 10.2 .14 1.30 6.0 2.8 125 meters wide. 26. ______53.80 16.3 .15 1.30 6.5 3.7 Aug. 2 ______56.20 12.3 .14 1.40 8.0 3.3 TABLE 6. Tonnage and analyses (in percent) of canga and en­ 9. __. 57.60 9.9 .12 L10 7.0 4.4 riched itabirite ore from the Cabral mine 16. ______59.10 8.0 .14 1.30 6.0 4.5 23 ______. _ 58.80 9.2 .16 1.70 7.0 2.8 [Tonnage figures and analyses were recorded by the Companhia Siderurgica Belgo- 30 ______58.80 7.0 .16 1.30 5.0 5.8 Mineira] Sept. 6.. .. 59.30 8.3 .10 1.00 4.0 3.7 9 ___ . _____ 58.80 8.7 .18 1.30 4.5 3.6 Metric Number 20 ______59.30 7.3 .12 1.40 5.0 4.3 Year tons of Fe 8iO2 P Mn AUOj Loss on 27 ____ . ____ 57.80 9.5 .20 1.10 4.0 4.5 shipped samples ignition Oct. 10 ______59.10 8.4 .16 1.30 3.5 4.5 analyzed .10 1.20 4.5 2.6 25.. 59.30 7.0 .25 1.20 4.0 4.8 1947...... 59,798 11 59.23 6.50 0.16 0.31 1.51 6.20 Nov. 3 ______60.10 7.2 .20 1.40 6.0 4.0 1948. _____ 31,890 12 59.78 6.06 .16 .20 1.46 6.06 II.-...... 57.09 10.8 .10 1.10 6.0 3.9 1949 ______20,800 12 58.40 9.10 .12 .14 1.53 9.10 1950 ______9,300 8 56.25 9.11 .17 .60 2.00 6.38 Average- _____ 58.05 9.35 0.12 1.22 5.5 3.9 1951. _____ 10,920 6 58.05 & 67 .17 .77 1.90 6.65 1952... ____ 10,200 11 57.76 & 44 .18 .79 1.85 6.44 1953..- ... 6,330 _ . - TABLE 5. Annual production j from the Gongo Sdco iron mine Average . . 58.25 7.31 .16 .47 1.71 6.80 [Records of the Companhia Ferro Brasileira, Caetg] Total 149, 238 ______. __ . ______...... Year Metric trnia Year Metric tons 1937 » ______956 1948 _ ___-_-__-._ 45,324 1938 _ .. ___ . _ 13,896 1949 ______32,615 MINA DA TRINDADE 1939 ______29,445 1950__-_-_- _ _. 8, 179 1940. ___ . _ _._- 43,880 1951 ______9,108 The Mina da Trindade is near N. 7,700, E. 11,000 (pi. 1941 ______.. 52,213 1952 ______11,636 2). The mine is on the south slopes of the Serra Geral, 1942 __ . ___ _.__ 52, 128 1953______. _ _ 15,771 north of the Central do Brasil railroad. It was operated 1943 __ .__. __ ._ 57,183 1954______25,765 by the Companhia Siderurgica Belgo-Mineira in 1952 1944. ______72,207 1955. ______21,626 and 1953, and approximately 80,000 tons of canga and 1945 ___ .. ___ __ 55,619 19562 ______19,216 1946______. 50,004 Total.... _ 669,863 enriched itabirite was mined. Company records show 1947. ______53,091 that monthly samples of the ore from the storage bins i December only. averaged 57.83 percent iron, 8.29 percent SiO2, 1.78 per­ ' January-October. cent A12O3, and 0.14 percent phosphorous. BARRA MANSA MINE The mine area is 350 meters long and 150 meters wide. The Barra Mansa mine is about 1,500 meters east of The ore, composed of surficial layers of canga overlying the Gongo ,S6co iron mine at N. 6,900, E. 8,600 (pi. 2). a thin layer of enriched itabirite, is similar to the ores The mine is worked as a series of benches cut into the exposed in the other mines along the south slopes of the south slope of the Serra Geral in an area 500 meters long Serra Geral. and about 100 meters wide. Production at the time of FURA OLHO MINE mapping (1957) was about 20 tons per day. The ore, The Fura Olho mine is at N. 5,900, E. 12,850 (pi. 2) . composed of canga and enriched itabirite, is shipped to the Barra Mansa smelter in the State of Rio de Janeiro. It was worked as a series of low benches cut into a canga capping. No production figures are available. The canga CABRAL MINE overlies itabirite, and the ore is composed of canga made The Cabral mine (Mina do Cabral) is north of the up of fragments of itabirite and some fragments of Central do Brasil railroad at the small village of Hen- hard hematite cemented with hydrous iron oxides. rique Fleuisse, near N. 7,350, E. 9,700 (pi. 2) . The mine area is on the south slopes of the Serra Geral and can PONTE DA PAIXAO MINE be reached by a road leading south from the main road The Ponte da Paixao mine is in the eastern part of leading east from Belo Horizonte through Sahara and the Gongo Soco quadrangle at N. 5,250, E. 12,000 (pi. 2) . ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 143 The mine can be reached by a road leading southwest (pi. 2). The mine area can be reached by a road leading from Barao de Cocais to the Rio Conceigao valley. The south from the village of Joao Vasconcelos to the Gongo mine is operated by the Companhia Brasileira da Usi- Soco gold mining area. nas Metalurgicas. Production started in June 1956, and Gold ores were mined from the Luis Scares mine dur­ at the time of this survey (1957), 6,000 tons of ore ing the early part of the 19th century, but there is no averaging 55.3 percent iron had been mined. The mine record of production. The quartz veins of the mine area is opened by three benches cut into the north east slopes were again prospected in 1945 by the Departamento of Alto Fernandez. Nacional da Produgao Mineral of the Brazilian govern­ The canga layer, 1-2 meters thick, is composed of ore ment. Core drilling along the crest of the Serra Geral ranging from poorly cemented chips of enriched itabi- intersected an auriferous quartz vein striking N. 85° E. rite and some hard hematite fragments to porous honey­ and dipping about 45° S. The vein was explored at two combed canga. levels by three adits driven into the north slopes of the Serra Geral. Approximately 600 meters of underground OTHER MINERAL RESOURCES workings was completed. The veins and wallrock were In addition to iron, mineral resources of the mapped too low in gold content to be economically mined, and area include gold, manganese, dolomite, and bauxite. the prospect was abandoned. At the time of the survey Iron, manganese, and dolomite have been mined for all workings were either collapsed or flooded. more than 20 years for use in the local iron smelters. Although the vein prospected in the Luiz Soares Gold was mined from placer and lode deposits during mine is poorly exposed at the surface in the mine area, the latter part of the 18th century and the early part of it is exposed in a roadcut about 300 meters west of the the 19th century by the Portuguese and British settlers. old working at N. 5,800, E. 1,200 (pi. 2). The vein Aluminous lateritic soils in the Lagoa Grande Plateau strikes west and dips 40°-45° S.; it occurs along the have not been mined but are a potential source of low- thrust plane of the Cambotas fault. The vein is strongly grade aluminum ore. sheared and contains limonite along the shear planes. Quartz stringers cut the dolomitic phyllite and quartz GOLD DEPOSITS schist; they also cut dolomitic itabirite in the upper Gold was obtained in the mapped area as early as plate and the quartz-chlorite schist of the lower plate. 1599 by the Indians, who used it to barter with early set­ Informal reports by the Departmento Nacional da tlers, but it was not until 1695 that gold was discovered Produgao Mineral state that the veins consist of milky in bedrock at the headwaters of the Corrego do Gongo, vein quartz, a few veinlets and clusters of pyrite crys­ near N. 5,000, E. 4,000 (pi. 2), in the Gongo Soco quad­ tals, sericite, and rare folia of metallic gold. rangle (Henwood, 1871, p. 360). From 1695 to the mid­ CAMARA MINE dle of the 18th century, gold placers were mined The Camara mine is at N. 5,950, E. 5,250, in the extensively along the Rio Socorro and the Rio Con­ Gongo Soco quadrangle (pi. 2). The mine was appar­ ceigao ; the remnants of extensive placer washings can ently developed during the early part of the 19th cen­ be seen along these major drainages across the mapped tury, and the trenches, open pits, and waste dumps are area. Records of placer mining operations are not now overgrown by bushes and trees. The mine was pros­ Jmown to the writer, and it is likely that few reports pected by a series of west-trending open pits and were ever prepared. Records of the earliest under­ trenches that were excavated across and along a sheared ground mining ventures are scant, but according to discordant contact between overlying itabirite and Burton (1860), in his description of the highlands of quartz schist. The contact is exposed over an area 600 Brazil, Coronel Manoel da Camara de Noronha inter­ meters long on the north slopes of a hill underlain by mittently mined the lode deposits of the Gongo Soco Caue Itabirite. The mine was developed by an inclined mine, at the headwaters of the C6rrego do Gongo, from shaft on the sheared contact that dips 40°-45° S. There about 1800 to 1808. Systematic mining of these lode de­ is no record of production, but a large waste dump at posits of the Gongo Soco gold mine was started in 1826 the western part of the mine area shows that there was by Captain Lyon of the Imperial Mining Association, considerable underground exploration of the quartz and gold was produced until 1856, when the mine was schist and itabirite. closed because of depletion of ore. The quartz schist of Nova Lima age beneath the LUf S SOARES MINE shear zone is cut by numerous tear fractures filled by The Luis Scares mine, which is now abandoned, is stringers of milky (bull) quartz. The laminations in the along the north slopes of the Serra Geral (N. 6,000, E. itabirite are deformed by complex drag folds. The itabi­ 1,300) in the western part of the Gongo Soco quadrangle rite is dragged into the shear zone and obliquely trun- 144 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL cated by it. The shear zone is well exposed in the partly diminishing grade of ore (Henwood, 1871, p. 331). To­ collapsed collar of the inclined shaft; it is about 50 cm tal production during this period by the Imperial Min­ wide and is made up mostly of granular quartz and ing Association was 34,427 pounds of gold. About 23,381 itabirite cut by stringers of quartz. pounds was produced by screening and hand sorting of The quartz veins in the surface exposures contain the coarser nuggets, wires, and folia of metallic gold, minor amounts of sericite and clusters of hematite crys­ and about 11,146 pounds of fine gold was recovered from tals, and some of the fractures in the quartz schist con­ stamp mills (Henwood, 1871). tain earthy dark-brown limonite and brownish-black According to Henwood (1871) the wallrock of the manganese oxide. The rocks of the waste dump are mine was generally divided into the "Gongo Soco series" mostly banded brown and white quartz schist cut by in the northern part of the mine workings and the narrow quartz veinlets and stringers containing sericite "Cunha series" in the southern part of the mine area. and specular hematite. Henwood (1871, p. 182) de­ The grade of the ore in the "Gongo Soco series" was scribed the gold mineralization as follows: richer and the ore shoots were more persistent in depth Quartz and earthy brown iron ore are thinly spotted with gold and strike length; most of the production came from of the finest quality in many isolated beds which from a mere rich pockets of metallic gold in the "Gongo Soco series." line perhaps to three feet wide, and from a few feet to eight Correlation of Henwood's surface sketch maps of the or ten fathoms long occur throughout the formation and mine area with the surface (pi. 2) shows the mine area accommodate themselves to all its undulations. to be traversed by the contact between the dolomite of Henwood's brief description does not indicate whether the Gandarela Formation and the Caue Itabirite. Hen- the gold occurs in the itabirite above or in the quartz wood's "Cunha series" apparently correlates with lower schist below the shear zone that separates them; how­ dolomite beds of the Gandarela Formation and his ever, the undulations he described may refer to the drag "Gongo Soco series" with the dolomitic itabirite in the folds in the itabirite. upper part of the Caue Itabirite. Apparently most of CONGO SdCO (GOLD) MINE the richer "jacutinga" ores of the "Gongo Soco series" The Gongo Soco gold mine was abandoned in 1856, came from the gradational contact zone between the and only ruins of the buildings and old mine workings dolomite and the dolomitic itabirite in the upper part can be seen. The mine area, about 2 kilometers long, is of the Caue Itabirite, and the lower grade ores of the at the base of the Serra Geral in the general vicinity of "Cunha series" apparently occurred in lower dolomite N. 4,800, E. 4,500 (pi. 2). The mining area can be beds of the Gandarela Formation. reached by a road that leads westward from the town of Henwood described a number of structural features Barao de Cocais along the Rio Socorro valley through that were associated with the richer ore shoots in the the village of Venda Velha. "Cunha" and "Gongo Soco" series. He noted (1871, p. The Gongo Soco mine was developed by Coronel 270-71) that the richer ore shoots in both series occurred Manoel da Camara de Noronha from about 1800 to 1808, in steeply dipping beds, whereas the lower grade ore when small tonnages of ore were mined from surface shoots occurred in shallow or flattened areas of the wall- and shallow underground workings. At his death the rock. He stated (p. 259) : "Near the richer (shoots- mine was willed to his son Isador, who sold the mine to bunches) masses of ore, certain strata exhibit broad Capitao Mor Alves da Cunha. Captain Cunha operated undulation (fig. 22, A-A'}." Henwood's figure 22 and the mine intermittently from 1808 to 1824 and then sold section A-A' show asymmetrical folds with numerous it to the British Imperial Mining Association for small-scale drag folds superimposed on the limbs of the £70,000 (Burton, 1860). asymmetrical folds. The Imperial Mining Association started operations The gold occurred mostly in lenses of "jacutinga," in 1826, and the mine was developed by a series of ver­ and Henwood (1871, p. 271) stated: "The richest bod­ tical shafts and crosscuts. A surface sketch map of the ies of jacutinga are seldom more than a few feet in mine workings and a cross section of the underground length and depth." He also stated (p. 259) : "Wherever workings by Henwood show the mine workings along the base and lower slopes of the Serra Geral in an area hard crystalline, massive iron-glance [hard hema­ approximately 1,500 meters long by about 400 meters tite^)] touches or approaches the auriferous bands, wide. According to Henwood (1871, p. 250), "the north­ they cease to be productive." He further noted (p. 341) ern or Gongo Soco deposits has been laid open by scores that both itabirite and "jacutinga" are traversed by of shafts and miles of levels." cross veins of quartz that seldom exceed a fathom (6 The Imperial Mining Association produced gold from feet) in length or depth and are less than an inch in 1826 to 1856 but then closed the mine because of the width. ANTONIO DOS SANTOS, CONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 145 Henwood's (p. 271-274) description of the gold min­ is not known; however, judging from Hussak's and eralization in the "Gongo Soco series" is as follows: Henwood's descriptions, "jacutinga" consisted mostly Near the middle of each bed the matrix, already described, of schistose hematite or soft hematite and may have encloses rough (nuggets) lumps, usually but a few ounces, been derived from the hard hematite. If this assump­ though here and there two or three pounds, in an instance or tion is valid, the question still remains whether the hard two of still greater weight. Some of these are isolated; others "iron glance" was softened by ground waters that en­ at unequal intervals are either irregularly clustered, or united riched the "jacutinga" or by hydrothermal fluids that by thin laminae and reticulating threads; but all are gold. Nuggets are common near the centers of aggregation; f oliae to­ precipitated gold in the "jacutinga." wards their circumferences. But whilst clusters, such as these, The occurrence of the higher grade ore at or near form a striking characteristic, smaller nuggets (prills), the surface suggests supergene enrichment of the ores flakes, threads and granules, slightly coherent or, more fre­ of the Gongo Soco mine. Henwood stated (p. 331): quently isolated in the surrounding matrix furnish the chief "Thus, at progressively greater depths, the productive riches of this formation. Grain particles and, sometimes, small nuggets occur in numbers gradually diminishing toward the beds so far as they have been wrought continue to confines of every bunch; on both sides of large masses however, yield smaller averages of gold." Dorr and Barbosa adjoining strata are for short distances thinly sprinkled with (1963, p. 105) described gold in the "jacutinga" of the gold. Itabira district as follows: "The concentration of gold The term "jacutinga" has been used for describing into incredibly rich pockets and seams is most reason­ soft itabirite and soft high-grade hematite ore within ably attributed to secondary enrichment by through- the region of the Quadrilatero Ferrifero. It is a local passing supergene waters, since the distribution of the term that has no clear scientific use; however, Henwood ore was related to the present erosion surface." Dorr and apparently used it to describe either partly disaggre­ Barbosa further suggested that manganese oxide in the gated schistose high-grade hematite or soft high-grade weathered dolomitic iron-formation, which is in a very hematite layers, rather than soft itabirite, because he finely divided state and therefore chemically active, noted that "jacutinga" was the most important host rock serves as a precipitating agent in the "jacutinga" ores. for gold mineralization, whereas itabirite was an unfa­ The reason for the diminution of grade with depth is vorable host rock. Henwood (1871) stated: not known. The diminution could be related to surface weathering and supergene enrichment of the dolomitic Itabirite consists for the most part of granular quartz and iron glance [specularite] irregularly mixed with oxydulated itabirite, as suggested by Dorr and Barbosa (1963, p. iron [magnetite crystals (?) ] earthy brown iron ore, and hydrous 105), or it could be due to lack of suitable host rock, oxide of iron in alternating 'beds, some times separated by lami­ namely dolomitic itabirite, for the precipitation of gold nae of talc * * * in this part of the series gold is rarely found from through-going hydrothermal fluid. At a depth of * * *. Jacutinga is composed in a great measure of iron-glance 150 meters, in the northern part of the mine area, the [soft hematite(?)] mixed generally with small quantities of earthy manganese and frequently with minute proportions of lower-most mine workings were probably in the lower oxydulated iron, earthy iron-ore, titaniferous iron ore, or hy­ beds of the Caue Itabirite, which are in most areas sili­ drous oxide of iron also, but seldom with all these at once. ceous itabirite, whereas the itabirite in general becomes Hussak (1908) described fragments of gold-bearing increasingly more dolomitic upward and grades through "jacutinga" from the Gongo Soco mine as "schistose itabiritic dolomite into dolomite of the Gandarela hematite, containing very little pyrolusite, earthy limo- Formation. The origin of the gold is not known, but a possible nite, some scales of talc, and certain masses which had the appearance of kaolinite." Hussak's (1908) and Hen- source could be gold-bearing hydrothermal solutions wood's (1871) descriptions of "jacutinga" are very emanating from the batholithic body of granitic gneiss similar, and it seems likely that the "jacutinga" of the that forms the Caete Complex. Gongo Soco mine was softened or partly softened high- MANGANESE DEPOSITS grade hematite lenses interlayered in the itabirite and Manganese ores have been mined in the Gongo Soco dolomite of the Itabira Group. quadrangle from the Capim Gordura and Lagoa das Henwood's description of the gold mineralization Antas mines since 1940 for use in the charcoal blast fur­ along drag folds cut by quartz veins in the Gongo Soco naces of the Companhia Brasileira da Usinas Metalur- mine seems to indicate a hydrothermal origin for the gicas in Barao de Cocais. The total production of man­ gold. Henwood observed (p. 329) that in all parts of ganese ore is not known, but it may be about 1,000 tons. the mine, hard, crystalline, massive iron-glance (hard The manganese ore occurs almost exclusively within hematite) seems to have an unfavorable influence upon the dolomite of the Gandarela Formation as supergene the grade of the gold in the ore shoots. The relation of concentrations of manganese oxide generally parallel the hard iron-glance to the gold-bearing "jacutinga" to the bedding planes of the dolomite. The dolomite con-

309-760 6E 146 GEOLOGY AND MINERAL RESOURCES OF PARTS OF MINAS GERAIS, BRAZIL tains varying amounts of manganese ranging from a LAG6A DAS ANTAS MINE trace to 1.2 percent, and upon weathering the indigenous The Lagoa das Antas anine is on the crest of the manganese of the dolomite is apparently concentrated Serra da Pedra Formosa (N. 3,700, E. 8,850, pi. 2) in by meteoric waters along bedding planes and shear zones the southwestern part of the Gongo Soco quadrangle in the dolomite (Dorr and others, 1956). and is named for a nearby lake. A road leads south­ eastward from the Kio Acima-Conceic.ao road along the C\PIM GORDURA MINE crest of the Serra da Pedra Formosa to the mine area. The Capim Gordura mine is in the western part of The prospect was developed by the Companhia the Gongo Soco quadrangle (N. 3,650, E. 1,080, pi. 2) at Brasileira da Usinas Metalurgicas of Barao de Cocais. the foot of the south slopes of the Serra Geral. The The mine consists of one open pit trending 1ST. 55° E. mine area is accessible by a road leading west from the that is 50 meters long, 25 meters wide, and 3-5 meters town of Barao de Cocais along the Socorro valley. deep. No production is recorded, and the grade and the The Capim Gordura mine is owned and operated by potential tonnage of available ore probably are not the Companhia Brasileira da Usinas Metalurgicas of high enough to warrant commercial mining. Barao de Cocais, and manganese has been mined inter­ The manganese minerals occur along a shear zone in mittently since 1940 for use in their charcoal blast the lower 50 meters of dolomite of the Gandarela furnaces. No accurate data of total production of manga­ Formation. This shear zone trends 1ST. 50° E. and is nese ore from the Capim Gordura mine have been re­ vertical. corded, but it is estimated that approximately 1,000 tons The manganese minerals in the shear zone occur as have been mined. Three pits were being mined at the dark-gray, sooty material and nodules and a few silver- time of the survey, but there are numerous caved and gray crystals of manganese oxide in highly sheared red abandoned pits. The ore is mined by hand tools, and it saprolite derived from dolomite of the Gandarela For­ is hand sorted to increase the grade of the ore. The re­ mation. An analysis of a channel sample cut across the serves are small, and the lenses and stringers of man­ 1-meter-wide mineralized shear zone shows that the ganese oxide are discontinuous in most pits and cannot sample contains 19.0 percent manganese (table 7). be projected laterally or vertically, with confidence, for more than a few meters. Prospecting for new lenses of DOLOMITE ore is costly, and the outlook for finding additional re­ Dolomite has been mined sporadically from small serves is not encouraging. quarries in the Socorro valley by the Companhia Brasi­ The manganese ore of the Capim Gordura workings leira da Usinas Metalurgicas of Barao de Cocais for occurs in saprolite derived from dolomite at horizons use in their charcoal blast furnaces. Most of the dolo­ mite is not desirable for use as flux because of the high ranging from 100 to 300 meters above the base of the MgO content, and the dolomite forms a highly viscous Gandarela Formation. In most places in the mine area slag that is difficult to remove from the furnaces. Anal­ the dolomite is weathered to red and brownish-black yses of fresh blocks of dolomite from the quarries of saprolite, although low ledges of fresh dolomite crop the Socorro valley are given in table 8. out in a few places. The ore consists of pods, lenses, and Hogbacks of dolomite exposed in Corrego da Gan­ stringers of manganese oxide ranging from about 1 cm darela valley near the west boundary of the Conceic.ao to 1 meter across. They are crudely alined parallel to do Rio Acima quadrangle (1ST. 7,000, E. 1,000, pi. 3) con­ the bedding planes of the saprolite, strike west, and dip sist of massive banded, red and white, white and gray, 40°-45° S. and cream-colored rock that could be quarried for orna­ The ore consists of either massive reddish-gray very mental stone or dimension stone. fine grained manganese oxide containing disseminated FEEETJGINOTJS BAUXITE clusters of hematite and quartz crystals and some limonite or vuggy botryoidal light-gray very fine Ferruginous bauxite soil derived from the weather­ ing of dolomite of the Gandarela Formation occurs as grained manganese oxide enclosing numerous silvery- irregular layers within the canga field of the Lagoa gray acicular crystals tenatively identified as manga- Grande Plateau in the northwestern part of the Con- nite. The vugs between the botryoidal growths of man­ ceigao do Rio Acima quadrangle. The ferruginous ganese oxide contain druses of acicular psilomelane bauxite grades laterally into hard chemical canga con­ crystals. sisting almost entirely of limonite. It is yellowish-red to Analyses of ore samples taken from the hand-sorted medium-red earthy material in which some layers are ore at the mine, and .analyses of the hand-sorted ore honeycombed by cellular cavities. It can easily be mined made by the company are shown in table 7. and may be a source of low-grade aluminum ore. ANTONIO DOS SANTOS, GONGO SOCO, AND CONCEIQAO DO RIO ACIMA QUADRANGLES 14?

TABLE 7. Analysis (in percent) of manganiferous laterite from Lagda das Antas mine and manganese ore from Capim Gordura mine [Analyses 1, 2, 3, and 4 made by Departamento Nacional da Producao Mineral, Belo Horizonte; analyst, Dr. Cassio Pinto. Analyses 1387,1401,1400,1397,1392,1409, and 1020 made by Companhia Brasileira da Usinas Metalurgicas; analyst unknown]

Lagda das Antas Capim Gordura mine mine

1387 1401 1400 1397 1392 1409 1920

SiO2-______-_-_.._..______. 3.5 6.6 2.6 17.5 12. 7 9. 48 10. 22 9. 74 7. 16 11. 86 10.44 Al_08______.______4.5 1. 2 1.2 1.3 _ Fe______-____..______--_ 37.2 7.4 19. 8 18.4 21.2 16. 0 17.4 16. 5 20.4 19. 0 18. 5 Mn_____. ______. _ ___ _ 19. 0 38.8 38. 5 30.4 40. 3 3Q 4 36.3 41. 3 39. 3 35. 1 40. 0 ?.__.____. ______._____.___. .06 .02 . 09 . 04 _

1. Manganiferous laterite taken from Lagoa das Antas mine at N. 4,000, E. 10,500 (pi. 2). Channel sample. 2-4. Manganese oxides taken from stockpile of ore at open pit at N. 3,700, E. 1,080 (pi. 2). 1387, 1401,1400,1397,1392,1409, and 1920 taken from ore bins at the smelter of Companhia Brasileira da Usinas Metalurgicas, Barao de Cocais.

TABLE 8. Analyses (in percent) of dolomite from the Gongo Soco quadrangle (pi. [Analyst: Dr. Cassio Pinto, Departamento Nacional da Producao Mineral, Belo Horizonte]

Sample Location Color CaO MgO Fe>O3 SiO2 A12O3 Mn s Loss on ignition

70____.__ N. 6700, E. 10,850..______White ______28.8 18. 6 2. 1 7.6 0.2 0. 2 0. 04 43. 2 7l_._____ N. 6700, E. 10,850. Gray 28. 7 18. 9 1.7 7.3 .2 . 2 .04 43. 2 72______N. 6700, E. 10,850. ___-___-______-_----do______29. 5 18. 5 2.3 4.7 .2 . 2 . 04 44. 6 73___._._ N. 6700, E. 10,850. __ _ _.______do______25.8 16.7 1.9 16.7 . 2 . 2 . 07 38. 7 74 ____ . N. 6600, E. 10,400._ __.______._. _ d(K______29.0 19.5 1.5 6. 1 . 2 . 2 .03 43. 7 75 __ ._. N. 6600, E. 10,400_ _. ______. ______do_. .__ __ 30.9 20.2 1.4 1.3 . 2 .2 .04 46. 1 76___.___ N. 6600, E. 10,400- ______--__-_------do ___ __. 29. 1 17. 2 1.8 9.7 .2 . 2 . 03 42. 1 84 ______N. 3000, E. 9600_ _ Cream______29.9 13.0 3. 1 15. 3 .3 . 2 . 02 37.9 84a __ ___ N. 3000, E. 9600__ _ . _ __ _ _ Gray. __ __ 24.7 17.3 2. 0 17.2 .5 . 2 . 02 38.3 87______N. 4950, E. 7900__ Cream__ _ _ _ 28.7 17.9 2. 0 8.6 . 3 . 2 .02 42.5 78______N. 6100, E. 9450______Gray____ 29.7 18.6 2.5 4. 7 .3 . 2 .05 44. 1 79_____-_ N. 6100, E. 9450__ . _ . . _ - . _____-do_-___-___ 26. 0 16. 5 3.6 12. 7 2. 2 .2 .04 38. 6 81______N. 3600, E. 4600- ______Red------30.3 20.8 .9 . 15 .3 .9 .03 46. 5 82_____._ N. 3600, E. 4600. ______do______- 30.4 20.7 . 8 . 15 . 1 . 1 .02 46.6 82g__.___ N. 3200, E. 2800_ ____..- ____-___-do______-_- 31. 0 19.9 1.5 . 10 . 3 .9 .02 46. 2 83__ _ ._ N. 3200, E. 2800__ _-_-______-_-__ -do______29.6 20. 1 3.6 . 15 .4 1. 2 .02 44.8 86 ______N. 1040, E. 5950- ______.__. _____-do______27. 6 17. 7 5. 0 6. 8 . 4 . 7 .02 41. 8

The thickness and horizontal extent of the exposures in thickness, but in most exposures along readouts and of ferruginous bauxite are not known because of a thin railroad cuts they average about 2 meters thick. The soil cover and thick overgrowth of scrubby jungle. A powdered specimens were tentatively identified in oils sample cut across a 30-meter traverse of ferruginous as kaolin clay. The color of the clay ranges from white bauxite along the trail southwest of Lagoa Grande through yellowish white, yellow, and reddish yellow contained 35.2 percent A12O3, 26.4 percent iron, 3.5 per­ to red. The white clay may be a source of material for cent SiO2, and traces of manganese, phosphorus, and refractory brick, and the yellow to red clay is suitable sulfur. for firebrick and roofing tile. CLAY None of the kaolin clay deposits have been prospected within the mapped area, but similar clay deposits have Clay deposits occur in a number of places as a sur- been mined for many years west of the mapped area ficial layer of lateritic soil mantling granitic gneiss near the city of Caete, where the clay is used in the saprolite in the Caete Complex. The residual clay soils manufacture of firebricks, refractory bricks, and roof range from a few centimeters to as much as 10 meters tiles. 148 GEOLOGY AND MINERAL RESOURCES Of PARTS OF MtNAS GERAIS, BRAZIL

REFERENCES CITED Barbosa, Octavio, 1954, Evolution du g^osynclinal Espinhago: Guimaraes, Djalma, 1951, Arqui-Brazil e sua evolugao geologica: Internat. Geol. Gong., 19th, Algiers 1952, Comptes rendus, Brazil, Div. Fomento da Produgao Mineral Bol. 88, 315 p. sec. 13, pt. 2 (fasc. 14), p. 17-36. 1958, Geologia estratigrafica e econSmica do Brazil: Bolau, E., 1952, Geology of the Caete"-uplift, central Minas Belo Horizonte, Brazil, Graficos Santa Maria S. A., 430 p. Gerais, Brazil: Geol. Foren. Stockholm Forh., v. 74, no. 4, Harder, E. C., and Chamberlin, R. T., 1915, The geology of cen­ p. 475-493. tral Minas Gerais, Brazil: Jour. Geology, v. 23, no. 4, p. Burton, Capt Richard F., 1860, Explorations of the highlands of 341-378, no. 5, p. 385-424. the Brazil: London, Tinsley Brothers. Henwood, W. J., 1871, Observations on metalliferous deposits: Derby, O. A., 1906, The Serra do Espinhago: Jour. Geology, v. Royal Geol. Soc. Cornwall Trans., v. 8, pt 1, 722 p., 14, no. 5, p. 374-401. Penzance. Dorr, J. V. N., 2d, 1958a, The CauS Itabirite: Soc. Brasileira Herz, Norman, and Dutra, C. V., 1958, Preliminary spectrochem- Geologia Bol., v. 7, no. 2, p. 61-62. ical age determination results on some granitic rocks of the 1958b, The Gandarela Formation: Soc. Brasileira Geolo­ Quadrilatero Ferrifero, Minas Gerais, Brazil: Soc. Brasi­ gia Bol., v. 7, no. 2, p. 63-64. leira Geologia Bol., v. 7, no. 2, p. 81-95. Dorr, J. V. N., 2d, and Barbosa, A. L. M., 1963, Geology and ore Hussak, Eugenio, 1908, Occurrence of palladium in Brazil: deposits of the Itabira district, Minas Gerais, Brazil: U.S. Mining Jour., London. Geol. Survey Prof. Paper 341-C, 110 p. Maxwell, C. H., 1958, The Batatal Formation: Soc. Brasileira Dorr, J. V. N., 2d, Coelho, I. S., and Horen, Arthur, 1956, The Geologia Bol., v. 7, no. 2, p. 60-61. manganese deposits of Minas Gerais, Brazil: Internat. Oliveira, A. I. de, 1956, Brazil, in Jenks, W. F., ed., Handbook Geol. Gong., 20th, Mexico City 1956, Symposium sobre of South American geology: Geol. Soc. America Mem. 65, yacimientos de manganese, v. 3, p. 279-346. p. 1-62. Dorr, J. V. N., 2d, Gair, J. E., Pomerene, J. B., and Rynearson, Pomerene, J. B., 1958a, The Cercadinho Formation: Soc. Brasi­ G. A., 1957, Revisao da estratigrafia Precambriana do leira Geologia Bol., v. 7, no. 2, p. 64-65. Quadrilatero Ferrif ero, Minas Gerais, Brasil: Brazil, Dept. 1958b, The Tabooes Quartzite: Soc. Brasileira Geologia Nac. Produgao Mineral, Div. Fomento Produgao Mineral Bol., v. 7, no. 2, p. 66-67. Avulso 81, 31 p. 1958c, The Barreiro Formation: Soc. Brasileira Geologia Dorr, J. V. N., 2d, Guild, P. W., and Barbosa, A. L. M., 1952, Bol., v. 7, no. 2, p. 67-68. Origin of Brazilian iron ores: Internat Geol. Gong., 19th, 1964, Geology and ore deposits of the Belo Horizonte, Algiers 1952. Symposium sur les gisements de fer du monde, Ibirite', and Macacos quadrangles, Minas Gerais, Brazil: v. 1, p. 286-298. U.S. Geol. Survey Prof. Paper 341-D, 84 p. Eschwege, W. L. von, 1822, Geognostisches Gemalde von Brasi- lien und wahrscheinliches Muttergestein der Diamanten: Rynearson, G. A., Pomerene, J. B., and Dorr, J. V. N., 2d, 1954, Weimar. Contacto basal de S6rie de Minas na parte ocidental do 1833, Pluto Brasiliensis: Berlin, G. Reimer, 622 p. Quadrilatero Ferrifero, Minas Gerais, Brasil: Brasil, Dept. Freyberg, Bruno von, 1932, Ergebnisse geologischer Forschun- Nac. ProdugSo Mineral, Div. Geologia, Avulso 34, 18 p. gen in Minas Geraes (Brasilien) : Neues Jahrb. Mineral ogie, Simmons, G. C., 1958, The FScho do Funil Formation: Soc. Geologic, u. Palaontologie, Sonderband 2, v. 2, 403 p. Brasileira Geologia Bol., v. 7, no. 2, p. 65-66. Gair, J. E., 1958, The Sahara Formation: Soc. Brasileira Geolo­ 1960, Origin of certain cangas of the Quadrilatero Fer­ gia Bol., v. 7, no. 2, p. 68-69. rifero of Minas Gerais, Brazil: Soc. Brasileira Geologia Bol., 1962, Geology and ore deposits of the Nova Lima and v. 9, no. 2, p. 38-59. Rio Acima quadrangles, Minas Gerais, Brazil: U.S. Geol. 1968, Geology and mineral resources of the Barao de Survey Prof. Paper 341-A, 67 p. Cocais area, Minas Gerais, Brazil: U.S. Geol. Survey Prof. Guild, P. W., 1957, Geology and mineral resources of the Congon- Paper 341-H (in press). has district, Minas Gerais, Brazil: U.S. Geol. Survey Prof. Sininions, G. C., and Maxwell, C. H., 1901, Grupo Tamandud da Paper 290, 90 p. Serie Rio das Velhas: Brazil, Div. Geologia e Mineralogia Guimaraes, Djalma, 1931, Contribuigao a geologia do Estado de Bol. 211, 31 p. Minas Gerais: Brazil, Servigo Geologia e Mineralogia Bol. Wallace, R. M., 1958, The Moeda Formation: Soc. Brasileira 55, 36 p. Geologia Bol., v. 7, no. 2, p. 59-60. INDEX

[Italic page numbers indicate major references] >

Canga, Barra Mansa mine Continued Page Contact relationship Continued Page Accessibility of report area. ______.. 12 Mina da Trhidade..------...... 142 Cau6 Itabirite and Gandarela Formation. 118 Acknowledgments_____.___..._.__ 3 ore grade__.....------. ..._ 41 Cua6 Itabirite and Nova Luna Group_. 18 AUuvium_...._.._---.__-----_-._-- 25 PontedaPaixgomine.-- - .___ 42 Cercadinho Formation and Sahara For­ Analyses, chemical, bauxite lateritic soil_ _ 25 total reserves._.-.. --..-. 41 mation...__..-______23 chemical, canga..--._____------42 types__.-______------__... 24 Maquing Group and Cambotas Quartzite. 9 dolomite__..__.____.____ 47 Capenema quadrangle, unconformities.___ 33 Moeda Formation and B atatal Formation. 15 Gandarela Formation.------..... 21 Capim Gordura mine, manganese.- .--.-.. 46 Moeda Formation and Nova Luna Group. 15 enriched itabirite, Cabral mine. .. -. -. 42 Capivarasyncline.------_------____. 29 Nova Luna Group and Batatal Forma­ Gongo Soco mine__--.-.. 42 Caraca fault____.___.-.. --__ SO tion...... 7 hard hematite ore, Cabeca de Ferro Caraga Group.______.. ____ 14,16 Nova Lima Group and Caet6 Complex_ 8 mine.-._.._...___...___ 39 Caraea Quartzite__...... __.__.._ 10,11,14 Nova Luna Group and Cambotas Quartz­ Piao prospect__._____,--._. 38 Catas Altas quadrangle, unconformities_ ... 33 ite . .. -..-. 11 Santo Antonio mine__------39 Cau6 Itabirite.------.------14,16 Nova Luna Group and Caufi Itabirite.... 7,11 manganese ore, Capim Gordura mine- 47 contact relations, with B atatal Formation. 16 Nova Lima Group and Maquing Group.. 8,9 manganiferous laterite, Lagoa das with Gandarela Formation....-...--- 18 Nova Lima Group and Minas Series_.-. 7 Antas mine__------47 with Nova Lima Group. ...._. 7,11,18 Nova Lima Group and Moeda Formation. 7 Argillite, Batatal Formation.__..-.------... 16 gold...... _-.._--. _ _ 39 Coordinate grid system, explanation__.-..-- 3 hard hematite ore.... _-.. _ ___ 36 Corrego do Gongo, gold deposits_------__ 43 Piao prospect.______. ___... 36 Crossbedding, Cambotas Quartzite_____ 12 B quartz-hematite veins__-..-..__..... 28 Cunha series....___.._._____------44 Barbacena Series.------..------4 Santo Antonio mine.- -... _._... 36 Barra Mansamine------.-.------.-_------3,42 soft hematite ore__----.--.---.---.---.. 39 Barreiro Formation-..------_ 14 D Central do Brasilrailroad-..---. .-__._ 2 Dikes, diabasic.-______..______25 Batatal Formation-...__-.---.._.__.. 14, IS Cercadinho Formation_...... _._... U,88 contact relations, with Cau6 Itabirite_. - 16 mafic..______27 contact relations with Sahara Formation.. 23 pegmatite.______26 with Moeda Formation.------15 quartz-kyanite veins.__._ _...-.-. 26,28 with Nova Lima Group..-__------7 Dolomite...... -.. . 3,46 Chapinlia__...... - _ .- .....__.. 17 analyses..___ . . ..._ 47 quartz-hematite veins. ______.__- 28 Chemical analyses, bauxite lateritic soil_ 25 Batatal Schist...... __._-.-----_.-_---_-.-.- 14 Gandarela Formation_____...____ 19 canga.______._... 42 chemical analyses_____._____ 21 Bauxite, ferrughious___------_.__. 46 dolomite____._ _ . ..-.-. 47 BeloHorizonte.------.------2 weathering products.__._____ 21 Gandarela Formation...... 21 manganiferous, Gandarela Formation...- 20 enriched itabirite, Cabral mine___.... 42 weathering products 25 Gongo Soco mine.. _._ 42 Cabeca de Ferro mine.-.__------3 hard hematite ore, Cabeca de Ferro mine- - 39 hard hematite ore-.-.------,-..- S8 Piao prospect_ -- 38 Santo Antonio mine-._-___----- 39 Fazenda Capivara.__._..._._._.__ 29 hydrothermal iron transfer ------20 Fficho do Funil Formation. ______14 Cabral mine.__._------_...___-.. 3,42 manganese ore, Capim Gordura mine.-.- 47 manganiferous laterite, Lagoa das Antas Feldspar, Caetg Complex.._____.___ 6 canga, analysis-..-.-.--- ..-. ------41 Fieldwork...... ______._____ 3 Caet6 Complex...------4,8,12,25 mine_ .-..._...... 47 Chert, Batatal Formation______. 16 Fundao fault.. ..-...... 31 contact relations, with Cambotas Quartz- Fura Olhomine, canga______42 ite .. ..-----. --.-- -. 13,26 Chlorite, Cercadinho Formation-__------23 with Nova Lima Group.___.___ 8,26 Chloritoid, Cercadinho Formation_____ 23 deuteric alteration______.____ 29 Clay deposits______. ______. 47 G mafic dikes....__---.--.---.-_------27 Climate_----__-____.___------3 Gandarela Formation._...______. U,19 Caetfi quadrangle, Nova Lima Group..-___ 4 Cocais quadrangle, age of granitic rocks__-. 27 Cabeea de Ferro mine_------__--... 36 unconformities------.-----.-----.---_ 33 unconformities.-._____._.- - 33 contact relation with Cau§ Itabirite...... 18 Camaramine, gold..__------_------43 Companhia Brasileira da Usinas Metalurgicas. 3, dolomite, chemical analyses...... __...... 21 Cambotas fault_-.-_-----.------.... si 38,43,45,46,47 weathering products. . .. 21 Cambotas homocline------..------.-- 33 Companhia Ferro Brasileira--___-___- 41,42 hard hematite ore..._-----____.... 30 Cambotas Quartzite_.-._.--_--..___ 4,9, 10 Companhia Ferro Brasileiro, Usina Gorceix.. 3 manganese ore______. ______45 contact relations, with Caet6 Complex_ 13,26 Companhia Siderurgica Belgo-Mineira_... 3,41,42 splash rock.._..--.-----..- -.. .. 21 with Maquinfi Group.-..-___.... 9 Conceicao anticline______... #9 Gandarela quadrangle, unconformities ------33 with Nova Lima Group- ______. 11,13 Conceigao valley..______-__ 29 Gandarela syncline.- ... 8, SO diabasic dikes.-----.----__------_. 25 schist and phyllite.______. 6 Gneiss-- -.-.__...--__.__..._._... mafic dikes...-.-_-__..__.-.-__. 27 Conglomerate--...... __.... 4 Gold.-.. ... quartz veins..-.--___._.____._ 28 Gandarela Formation-______20 Cau6 Itabirite.-.------.-..- weathering products__---__.._..._ 12 Moeda Formation______...... 15 Gongo Soco goldmine___.-.--..-.---- 44 Canga, BarraMansamine..-._.-____-.- 42 Contact relationship, Batatal Formation and Luis Scares mine..-._._____-..--.. 43 Cabral mine____.____-...-.-...... 42 Caue Itabirite....-.-_... . - 16 Gongo Soco gold mine...._.______3, 43,44 definition__-----_____...____ 23 Caetg Complex and Cambotas Quartzite. 26 Gongo Soco iron mine..--.____...... 3 Fura Olhomine______. _. _ 42 Caetfi Complex and Nova Luna Group_ 26 canga.------. 41 Gongo Soco mine______. ______41 Cambotas Quartzite and Caetg Complex. 13 soft hematite ore...-._._._ -...-.---.- 40 analyses______._._ 42 Cambotas Quartzite and Nova Lima Gongo Soco series 44 Iron content_------______35,41 Group_. -.--.__... 13 Granitic gneiss. _ .----- ...... 96 I 49 150 INDEX

Page Page Page Granodiorite gneiss, mafic dikes__. ___ _ 127 Landslide...... -----.- 125 Previous investigations...... 13 weathering products.-----. ---...- . 25 Lateritic soils_____.... 4,25 Psilomelane crystals, Capim Gordura mine 46 Granodlorltic gneiss..------_-.----.-.-.-. tB Limonite, phenocrysts_.-. ....-..-.. 27 Pyrite, Gandarela Formation_...... 20 pseudomorphs after pyrite. - 27 Location of report area .- - - 2 Luts Scares mine, gold - 43 Quartz crystals, fibrous, Cambotas Quartzite- 28 Hard hematite, Iron content...... 35 Quartz veins_ 26 Hard hematite ore, Cabeca de Ferro mine.. S8 M Quartzite, age - - - 10 39 Mafic dikes. . 27 Cercadinho Formation, high-grade..--.----.--. _. . . se Magnetite, Caufi Itabirite 17 ferruginous, definition.. PiSo prospect...... ____.___... S6 Gandarela Formation_._-. .- -.-.- 19 Gandarela Formation.. analysis,___...______. 38 Manganese.....-- ...... - 3 Moeda Formation. _.. Santo Antonio mine-...... --.--.-_-.... Capiin. Gorduramine.. ... - 46 Sabara Formation . analysis-..__------___..-.-.... Lagoa das Antas mine..-.. _ . 46 B Hard itabirite-..__...__.._____.. 40 Manganese ores__...--. . - 45,47 References cited 48 Hematite, Caetfi Complex. 6 Manganiferous laterite, Lagoa das Antas mine, Rio da ConceioeSo____ 3 Cambotas Quartzite ____...-...... 13 analyses._ - 47 gold deposits._ 43 Cau6 Itabirite-...... 17 Maquine' Group..__.... . 4,S Rio das Velhas ...... 3 Cercadinho Formation.... _-. .. 22 contact relation, with Cambotas Quartz­ Rio das Velhas Series...... 4,10,14 crystals, Camaramine...... 44 ite 9 metamorphism.- 28 Gandarela Formation.______..... 19 with Nova Lima Group..... 8,9 Rio Doce- - 3 ore, high grade, definition. ___...... 16 quartz-kyanite veins__ .-.. - 26,28 Rio Socorro__ . 3 pseudomorphs after pyrite_ .... 27 Metasedimentary rocks.- 4 gold deposits 43 specular, Batatal Formation______. 16 Mina da Trindade, canga__ - 42 Rivers in report area... 3 See also Hard hematite and Soft hematite. Minas Series... 4,10,14 Henrique Fleuisse, Cabralmine, Canga...... 42 contact relation with Nova Lima Group.. 7 3 Henwood, W. J., quoted __.....____. 44,45 metamorphism.------28 Sabara Formation..._...... 14,22,23 Mining history _. _ . - 3 , Cercadinho Formation-_____. 22 1,1 Moeda Formation.------10,11,14 Santa Barbara quadrangle, Gandarela Forma­ Igneous rocks.. ------...... 4,25 contact relation, with Batatal Formation. 15 tion, thickness__...... 21 Caetfi Complex, relation to Cambotas with Nova Lima Group ... 7,15 Tamandua Group_...______. 4 Quartzite ...... 13 unconformities.... 33 Imperial Mining Association- _____. ___ 43,44 Santo Antonio mine, hard hematite ore_ S9 N Imperial Mining Co___-_..-._-.__-.-._ 3 Saprolite, definition.__...__...... 6 Industries of report area.___-..______2 Nova Lima Group. - 4 Nova Lima Group...... 6 Iron...... 3 contact relation, with Batatal Formation. 7 Sabara Formation______23 Iron-formation._... . . _ ...... 4 with Caet6 Comples...... 8,26 Schist, Cercadinho Formation.. 22 Nova Lima Group..._...._...... _ 8 with Cambotas Quartzite___.. 11,13 Nova Lima Group_ . . .. 6 Iron-formation conglomerates, Cambotas with Cau6 Itabirite.. . . 7,11,18 quartz-sericite, Moeda Formation____ 16 Quartzite_...... ____...._ 13 withMaquinfi Group...... 8,9 Sabara Formation______23 Iron ore. . . SB with Minas Series__-....--._ 7 Sericite, Moeda Formation_ 15 Itabira Formation.._...... 14,19 with Moeda Formation...... 7,15 Sericite laminae, Cambotas Quartzite____ 13 Itabira Group.______14,16 iron-formation_..._ -... 6 Serra da Pedra Formosa. 3 Itabirite-.-...... ,....._...... 40 pegmatite dikes---. . - .. 26 Serradas Cambotas-- . 3,10,31 Cau6 Itabirite, chemical composition__ 18 quartz-tourmaline veins.. . . . 26 Serra do Caraca...... 3,30 definition. ____.....___... weathering products.. . 4 mafic dikes. _.______.__. 27 enriched, Barra Mansa mine___..... Nova Lima quadrangle. 11 Serra do Maquing. 3 Cabralmine______..... Nova Lima Group.. ----- 4 Serra Espinhaco. 10,31 analyses...__..______. Serra QeraL . . 3 Gongo Socomine__....____. Soft hematite ore, Cabeca de Ferro mine.. 38 Palmital Formation______8 high-grade...... ___.__.__. S9 Mina da Trindade______. Pedra fault...... , .... 32 PiSo prospect______38 Gandarela Formation..- ______. Pegmatite 26,28 Soft Itabirite. _...... __ 40 iron content_____..._..._...... Phlogopite, Gandarela Formation______20 Soils, lateritie....__...... _... 25 magnetic...______18 Phyllite, Batatal Formation.,....._____ 16 Specularite, hard hematite ore__.... . 37 manganiferous_____.______20 Cercadinho Formation._...._._.... 22 Splash rock, Gandarela Formation_____. 21 Itacoloml Quartzite______14 Gandarela Formation-....-. -... __ 19 Structural features, Interpretation_____.. SS Itacolomi Series...... 4,10,11,14 Sabara Formation___..______23 T weathering products..- - ___... 6 Tabooes Quartzite.______14 Jacutinga. 45 PiSo prospect, hard hematite ore-.....-.-.... S3 Talus __.___._.__ 25 gold... 9,44 Pinto, Cassio, analyst____....____. 21,39,47 Tamandua Group. ______. 4,10 Piracicaba Formation...__._____..... 14 Thrust faults.-... __.-..______. SO Piracicaba Group.. ..._...___ ....- 14,22 Tourmaline crystals...._....______. 26 Plagioclase metacrysts, Cambotas Quartzite-- 13 Trindade mine______3 Lag5a das Antas mine, manganese____... 46 Ponte da Paixao mine, canga..-..------42 Lag5a GrandePlateau.....______24 Precambrian rocks, correlation chart__..... 5 ferruginous bauxite__..______.-.-. 46 structural development...... SS Vegetation in report area.